Researcher profile

Jian Zhang

Jian Zhang contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author

109works

0followers

44topics

4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

$A^3$-Bench: Benchmarking Memory-Driven Scientific Reasoning via Anchor and Attractor Activation

Scientific reasoning relies not only on logical inference but also on activating prior knowledge and experiential structures. Memory can efficiently reuse knowledge and enhance reasoning consistency and stability. However, existing benchmarks mainly evaluate final answers or step-by-step coherence, overlooking the \textit{memory-driven} mechanisms that underlie human reasoning, which involves activating anchors and attractors, then integrating them into multi-step inference. To address this gap, we propose $A^3$-Bench~ https://a3-bench.github.io, a benchmark designed to evaluate scientific reasoning through dual-scale memory-driven activation, grounded in Anchor and Attractor Activation. First, we annotate 2,198 science reasoning problems across domains using the SAPM process(subject, anchor & attractor, problem, and memory developing). Second, we introduce a dual-scale memory evaluation framework utilizing anchors and attractors, along with the AAUI(Anchor--Attractor Utilization Index) metric to measure memory activation rates. Finally, through experiments with various base models and paradigms, we validate $A^3$-Bench and analyze how memory activation impacts reasoning performance, providing insights into memory-driven scientific reasoning.

preprint2026arXiv

A Simplex Witness Certificate for Constant Collapse in Variational Autoencoders

We study exact constant collapse in variational autoencoders, where the deterministic encoder path becomes independent of the input. The VAE prior is kept as the standard Gaussian. Before VAE training, we construct a single fixed teacher posterior by searching a GMM-based approximation of the data. We then attach a fixed latent-only simplex witness to the encoder mean and compare its output with the teacher. The resulting alignment loss has an exact constant-predictor baseline: if the latent witness beats this baseline, the encoder mean cannot be input-independent constant. The same construction also gives a closed-form latent target that realizes zero teacher-witness alignment error for any full-support teacher posterior. This yields a concrete design principle: choose a teacher with nontrivial information but controlled log-odds energy, fix the witness, train only the encoder and decoder, and certify non-collapse by a positive margin. We present the theory, a minimal training protocol, and preliminary MNIST sanity checks. The analysis targets exact constant collapse. Reconstruction quality, sampling quality, and other collapse modes are evaluated with additional diagnostics rather than folded into the certificate itself.

preprint2026arXiv

A Testable Certificate for Constant Collapse in Teacher-Guided VAEs

Posterior collapse in variational autoencoders is often diagnosed by its symptoms: a small KL term, a strong decoder, or weak use of the latent code. These signals are useful, but they do not define a collapse boundary. We study a concrete failure mode, input-independent constant collapse, and show that this case admits an exact threshold. For any fixed nonconstant teacher distribution $T(\cdot\mid x)$, the best constant student is the dataset-average teacher distribution, and its alignment cost is the teacher mutual information $I_T(X;T)$. Therefore, if a strictly latent-only raw witness achieves alignment loss below this value, with a safety margin, the witness cannot be constant in the input. This identity turns a qualitative failure mode into a measurable one. In CIFAR-100 experiments with per-seed teacher search, full training stays on the certified side of the boundary, removing alignment drives the raw witness into the constant-student regime, and restarting from a collapsed checkpoint with alignment enabled restores the certificate. Tiny-ImageNet-200 fixed-target runs show the same prevention--collapse--rescue pattern across three independently searched teachers. Standard VAE-style baselines, including methods that preserve reconstruction quality or post-hoc predictability, remain negative under the raw certificate. The guarantee is intentionally narrow: it certifies that the matched nonconstant teacher-relative variation passes through the latent pathway, rather than claiming that all forms of posterior collapse have been ruled out.

preprint2026arXiv

Are Tools Always Beneficial? Learning to Invoke Tools Adaptively for Dual-Mode Multimodal LLM Reasoning

Tool-augmented reasoning has emerged as a promising direction for enhancing the reasoning capabilities of multimodal large language models (MLLMs). However, existing studies mainly focus on enabling models to perform tool invocation, while neglecting the necessity of invoking tools. We argue that tool usage is not always beneficial, as redundant or inappropriate invocations largely increase reasoning overhead and even mislead model predictions. To address this issue, we introduce AutoTool, a model that adaptively decides whether to invoke tools according to the characteristics of each query. Within a reinforcement learning framework, we design an explicit dual-mode reasoning strategy with mode-specific reward functions to guide the model toward producing accurate responses. Moreover, to prevent premature bias toward a single reasoning mode, AutoTool jointly explores and balances tool-assisted and text-centric reasoning throughout training, and promotes free exploration in later stages. Extensive experiments demonstrate that AutoTool exhibits outstanding performance and high efficiency, yielding a 21.8\% accuracy gain on V* benchmark compared to the base model, and a 44.9\% improvement in efficiency over existing tool-augmented methods on POPE benchmark. Code is available at https://github.com/MQinghe/AutoTool.

preprint2026arXiv

Central elements and evaluation map for the quantum queer superalgebras

We consider the $R$-matrix presentations of the quantum queer superalgebra $U_q(q_n)$ and its affine counterpart $U_q(\widehat q_n)$. We derive crossing symmetry relations for the $R$-matrices and use them to construct central elements in both superalgebras. We also produce an epimorphism $ev:U_q(\widehat q_n)\to U_q(q_n)$ identical on the subalgebra isomorphic to $U_q(q_n)$.

preprint2026arXiv

DAST: Difficulty-Adaptive Slow-Thinking for Large Reasoning Models

Recent advancements in slow thinking reasoning models have shown exceptional performance in complex reasoning tasks. However, these models often exhibit overthinking (generating redundant reasoning steps for simple problems), leading to excessive computational resource usage. While current mitigation strategies uniformly reduce reasoning tokens, they risk degrading performance on challenging tasks that require extended reasoning. This paper introduces Difficulty-Adaptive Slow Thinking (DAST), a novel framework that enables models to autonomously adjust the length of Chain-of-Thought (CoT) based on problem difficulty. We first propose a Token Length Budget (TLB) metric to quantify difficulty, then leverage budget-aware reward shaping and budget preference optimization to implement DAST. DAST penalizes overlong responses for simple tasks while incentivizing sufficient reasoning for complex problems. Experiments on diverse datasets and model scales demonstrate that DAST effectively mitigates overthinking (reducing token usage by over 30\% on average) while preserving reasoning accuracy on complex problems. Our codes and models are available at https://github.com/AnonymousUser0520/AnonymousRepo01.

preprint2026arXiv

EditRefiner: A Human-Aligned Agentic Framework for Image Editing Refinement

Recent text-guided image editing (TIE) models have made remarkable progress, yet edited images still frequently suffer from fine-grained issues such as unnatural objects, lighting mismatch, and unexpected changes. Existing refinement approaches either rely on costly iterative regeneration or employ vision-language models (VLMs) with weak spatial grounding, often resulting in semantic drift and unreliable local corrections. To address these limitations, we first construct EditFHF-15K, a dataset of fine-grained human feedback for edited images, comprising (1) 15K images from 12 TIE models spanning 43 editing tasks, (2) 60K annotated artifact regions and 80K editing failure regions, each accompanied by textual reasoning, and (3) 45K mean opinion scores (MOSs) assessing perceptual quality, instruction following, and visual consistency. Based on EditFHF-15K, we propose EditRefiner, a hierarchical, interpretable, and human-aligned agentic framework that reformulates post-editing correction as a human-like perception-reasoning-action-evaluation loop. Specifically, we introduce: (1) a perception agent that detects contextual saliency maps of artifacts and editing failures, (2) a reasoning agent that interprets these perceptual cues to perform human-aligned diagnostic inference, (3) an action agent that uses the reasoning output to plan and execute localized re-editing, and (4) an evaluation agent that assesses the re-edited image and guides the action agent on whether further refinements are required. Extensive experiments demonstrate that EditRefiner consistently outperforms state-of-the-art methods in distortion localization, diagnose accuracy and human perception alignment, establishing a new paradigm for self-corrective and perceptually reliable image editing. The code is available at https://github.com/IntMeGroup/EditRefiner.

preprint2026arXiv

ErrEval: Error-Aware Evaluation for Question Generation through Explicit Diagnostics

Automatic Question Generation (QG) often produces outputs with critical defects, such as factual hallucinations and answer mismatches. However, existing evaluation methods, including LLM-based evaluators, mainly adopt a black-box and holistic paradigm without explicit error modeling, leading to the neglect of such defects and overestimation of question quality. To address this issue, we propose ErrEval, a flexible and Error-aware Evaluation framework that enhances QG evaluation through explicit error diagnostics. Specifically, ErrEval reformulates evaluation as a two-stage process of error diagnosis followed by informed scoring. At the first stage, a lightweight plug-and-play Error Identifier detects and categorizes common errors across structural, linguistic, and content-related aspects. These diagnostic signals are then incorporated as explicit evidence to guide LLM evaluators toward more fine-grained and grounded judgments. Extensive experiments on three benchmarks demonstrate the effectiveness of ErrEval, showing that incorporating explicit diagnostics improves alignment with human judgments. Further analyses confirm that ErrEval effectively mitigates the overestimation of low-quality questions.

preprint2026arXiv

GenShield: Unified Detection and Artifact Correction for AI-Generated Images

Diffusion-based image synthesis has made AI-generated images (AIGI) increasingly photorealistic, raising urgent concerns about authenticity in applications such as misinformation detection, digital forensics, and content moderation. Despite the substantial advances in AIGI detection, how to correct detected AI-generated images with visible artifacts and restore realistic appearance remains largely underexplored. Moreover, few existing work has established the connection between AIGI detection and artifact correction. To fill this gap, we propose GenShield, a unified autoregressive framework that jointly performs explainable AIGI detection and controllable artifact correction in a closed loop from diagnosis to restoration, revealing a mutually reinforcing relationship between these two tasks. We further introduce a Visual Chain-of-Thought based curriculum learning strategy that enables self-explained, multi-step ``diagnose-then-repair'' correction with an explicit stopping criterion. A high-quality dataset with large-scale ``artifact-restored'' pairs is also constructed alongside a unified evaluation pipeline. Extensive experiments on our correction benchmark and mainstream AIGI detection benchmarks demonstrate state-of-the-art performance and strong generalization of our method. The code is available at https://github.com/zhipeixu/GenShield.

preprint2026arXiv

KV-RM: Regularizing KV-Cache Movement for Static-Graph LLM Serving

Static-graph LLM decoders provide predictable launches, fixed tensor shapes, and low submission overhead, but online decoding exposes highly irregular KV-cache behavior: request lengths differ, EOS events arrive asynchronously, and logical histories fragment over time. Dynamic runtimes recover flexibility through paged KV management and step-level scheduling, while static-graph executors often over-reserve memory and suffer burst-time latency outliers. This paper studies whether much of this variability can be absorbed below a fixed decode interface. We present KV-RM, a runtime design that regularizes KV-cache movement beneath a static-graph LLM decoder. KV-RM decouples logical KV histories from physical storage, tracks active KV state through a block pager, and materializes each decode step through a single committed descriptor. A merge-staged transport path coalesces non-contiguous KV mappings into a small number of large transfer groups before a fixed-shape attention kernel consumes them. Optional bounded far-history summaries can be enabled under the same interface, but the core design does not depend on them. On a 2-GPU NVIDIA A100 node, KV-RM improves mixed-length decoding throughput and tail latency relative to a static-graph baseline, reduces reserved KV memory across workload families, and removes severe burst-time latency spikes under production-trace replay. These results suggest that KV-cache movement, rather than kernel shape, can be an effective boundary for recovering runtime flexibility in static-graph LLM serving.

preprint2026arXiv

Leveraging Flatness to Improve Information-Theoretic Generalization Bounds for SGD

Information-theoretic (IT) generalization bounds have been used to study the generalization of learning algorithms. These bounds are intrinsically data- and algorithm-dependent so that one can exploit the properties of data and algorithm to derive tighter bounds. However, we observe that although the flatness bias is crucial for SGD's generalization, these bounds fail to capture the improved generalization under better flatness and are also numerically loose. This is caused by the inadequate leverage of SGD's flatness bias in existing IT bounds. This paper derives a more flatness-leveraging IT bound for the flatness-favoring SGD. The bound indicates the learned models generalize better if the large-variance directions of the final weight covariance have small local curvatures in the loss landscape. Experiments on deep neural networks show our bound not only correctly reflects the better generalization when flatness is improved, but is also numerically much tighter. This is achieved by a flexible technique called "omniscient trajectory". When applied to Gradient Descent's minimax excess risk on convex-Lipschitz-Bounded problems, it improves representative IT bounds' $Ω(1)$ rates to $O(1/\sqrt{n})$. It also implies a by-pass of memorization-generalization trade-offs.

preprint2026arXiv

LLM-Guided Quantified SMT Solving over Uninterpreted Functions

Quantified formulas with Uninterpreted Functions (UFs) over non-linear real arithmetic pose fundamental challenges for Satisfiability Modulo Theories (SMT) solving. Traditional quantifier instantiation methods struggle because they lack semantic understanding of UF constraints, forcing them to search through unbounded solution spaces with limited guidance. We present AquaForte, a framework that leverages Large Language Models to provide semantic guidance for UF instantiation by generating instantiated candidates for function definitions that satisfy the constraints, thereby significantly reducing the search space and complexity for solvers. Our approach preprocesses formulas through constraint separation, uses structured prompts to extract mathematical reasoning from LLMs, and integrates the results with traditional SMT algorithms through adaptive instantiation. AquaForte maintains soundness through systematic validation: LLM-guided instantiations yielding SAT solve the original problem, while UNSAT results generate exclusion clauses for iterative refinement. Completeness is preserved by fallback to traditional solvers augmented with learned constraints. Experimental evaluation on SMT-COMP benchmarks demonstrates that AquaForte solves numerous instances where state-of-the-art solvers like Z3 and CVC5 timeout, with particular effectiveness on satisfiable formulas. Our work shows that LLMs can provide valuable mathematical intuition for symbolic reasoning, establishing a new paradigm for SMT constraint solving.

preprint2026arXiv

LLMdoctor: Token-Level Flow-Guided Preference Optimization for Efficient Test-Time Alignment of Large Language Models

Aligning Large Language Models (LLMs) with human preferences is critical, yet traditional fine-tuning methods are computationally expensive and inflexible. While test-time alignment offers a promising alternative, existing approaches often rely on distorted trajectory-level signals or inefficient sampling, fundamentally capping performance and failing to preserve the generative diversity of the base model. This paper introduces LLMdoctor, a novel framework for efficient test-time alignment that operates via a patient-doctor paradigm. It integrates token-level reward acquisition with token-level flow-guided preference optimization (TFPO) to steer a large, frozen patient LLM with a smaller, specialized doctor model. Unlike conventional methods that rely on trajectory-level rewards, LLMdoctor first extracts fine-grained, token-level preference signals from the patient model's behavioral variations. These signals then guide the training of the doctor model via TFPO, which establishes flow consistency across all subtrajectories, enabling precise token-by-token alignment while inherently preserving generation diversity. Extensive experiments demonstrate that LLMdoctor significantly outperforms existing test-time alignment methods and even surpasses the performance of full fine-tuning approaches like DPO.

preprint2026arXiv

MAXS: Meta-Adaptive Exploration with LLM Agents

Large Language Model (LLM) Agents exhibit inherent reasoning abilities through the collaboration of multiple tools. However, during agent inference, existing methods often suffer from (i) locally myopic generation, due to the absence of lookahead, and (ii) trajectory instability, where minor early errors can escalate into divergent reasoning paths. These issues make it difficult to balance global effectiveness and computational efficiency. To address these two issues, we propose meta-adaptive exploration with LLM agents https://github.com/exoskeletonzj/MAXS, a meta-adaptive reasoning framework based on LLM Agents that flexibly integrates tool execution and reasoning planning. MAXS employs a lookahead strategy to extend reasoning paths a few steps ahead, estimating the advantage value of tool usage, and combines step consistency variance and inter-step trend slopes to jointly select stable, consistent, and high-value reasoning steps. Additionally, we introduce a trajectory convergence mechanism that controls computational cost by halting further rollouts once path consistency is achieved, enabling a balance between resource efficiency and global effectiveness in multi-tool reasoning. We conduct extensive empirical studies across three base models (MiMo-VL-7B, Qwen2.5-VL-7B, Qwen2.5-VL-32B) and five datasets, demonstrating that MAXS consistently outperforms existing methods in both performance and inference efficiency. Further analysis confirms the effectiveness of our lookahead strategy and tool usage.

preprint2026arXiv

MemFine: Memory-Aware Fine-Grained Scheduling for MoE Training

The training of large-scale Mixture of Experts (MoE) models faces a critical memory bottleneck due to severe load imbalance caused by dynamic token routing. This imbalance leads to memory overflow on GPUs with limited capacity, constraining model scalability. Existing load balancing methods, which cap expert capacity, compromise model accuracy and fail on memory-constrained hardware. To address this, we propose MemFine, a memory-aware fine-grained scheduling framework for MoE training. MemFine decomposes the token distribution and expert computation into manageable chunks and employs a chunked recomputation strategy, dynamically optimized through a theoretical memory model to balance memory efficiency and throughput. Experiments demonstrate that MemFine reduces activation memory by 48.03% and improves throughput by 4.42% compared to full recomputation-based baselines, enabling stable large-scale MoE training on memory-limited GPUs.

preprint2026arXiv

ReAlign: Generalizable Image Forgery Detection via Reasoning-Aligned Representation

The rise of AI-generated images (AIGIs) poses growing challenges for digital authenticity, prompting the need for efficient, generalizable image forgery detection systems. Existing methods, whether non-LLM-based or LLM-based, exhibit distinct advantages and limitations. While non-LLM-based models offer efficient low-level artifact detection, they often lack semantic understanding. Conversely, LLM-based methods provide strong semantic reasoning and explainability but are computationally intensive and less sensitive to subtle visual artifacts. Moreover, the true contribution of explanatory reasoning texts to forgery detection performance remains unclear. In this work, we investigate the intrinsic value and potential of LLM-generated reasoning texts, considering it a source of generalization and semantic-error sensitivity. Based on these findings, we propose ReAlign, a novel framework that distills high-quality reasoning texts generated by a GRPO-optimized LLM into a lightweight AIGI detector via contrastive learning. ReAlign effectively inherits the generalization ability and semantic sensitivity capability of reasoning textual representations, while remaining efficient and lightweight for deployment. Moreover, ReAlign adopts a tailored joint optimization strategy that integrates contrastive loss for image-text alignment and classification loss for accurate forgery discrimination. Experimental results on AIGCDetectBenchmark, AIGI-Holmes, and our newly constructed UltraSynth-10k demonstrate that ReAlign consistently outperforms existing state-of-the-art detectors in both accuracy and generalization, particularly when facing complex, high-fidelity forgeries from modern generative models.

preprint2026arXiv

Spark3R: Asymmetric Token Reduction Makes Fast Feed-Forward 3D Reconstruction

Feed-forward 3D reconstruction models based on Vision Transformers can directly estimate scene geometry and camera poses from a small set of input images, but scaling them to video inputs with hundreds or thousands of frames remains challenging due to the quadratic cost of global attention layers. Recent token-merging methods accelerate these models by compressing the token sequence within the global attention layers, but they apply a uniform reduction to query tokens and key-value tokens, ignoring their functionally distinct roles in 3D reconstruction. In this work, we identify a key property of feed-forward 3D reconstruction models: query tokens encode view-specific geometric requests and are sensitive to compression, while key-value tokens represent shared scene context and tolerate aggressive compression. Guided by this insight, we propose Spark3R, a training-free acceleration framework that decouples the compression of query tokens and key-value tokens by assigning distinct reduction factors, with intra-group token merging applied to query tokens and lightweight token pruning to key-value tokens. Additionally, Spark3R adaptively adjusts the key-value reduction factor across layers, further improving the quality-efficiency trade-off. As a plug-and-play framework requiring no retraining, Spark3R integrates directly into multiple pretrained feed-forward 3D reconstruction models, including VGGT, $π^3$, Depth-Anything-3, and VGGT-$Ω$, and achieves up to $28\times$ speedup on 1,000-frame inputs while maintaining competitive reconstruction quality.

preprint2026arXiv

StableI2I: Spotting Unintended Changes in Image-to-Image Transition

In most real-world image-to-image (I2I) scenarios, existing evaluations primarily focus on instruction following and the perceptual quality or aesthetics of the generated images. However, they largely fail to assess whether the output image preserves the semantic correspondence and spatial structure of the input image. To address this limitation, we propose StableI2I, a unified and dynamic evaluation framework that explicitly measures content fidelity and pre--post consistency across a wide range of I2I tasks without requiring reference images, including image editing and image restoration. In addition, we construct StableI2I-Bench, a benchmark designed to systematically evaluate the accuracy of MLLMs on such fidelity and consistency assessment tasks. Extensive experimental results demonstrate that StableI2I provides accurate, fine-grained, and interpretable evaluations of content fidelity and consistency, with strong correlations to human subjective judgments. Our framework serves as a practical and reliable evaluation tool for diagnosing content consistency and benchmarking model performance in real-world I2I systems.

preprint2026arXiv

StableMind: Source-Free Cross-Subject fMRI Decoding with Regularized Adaptation

Existing cross-subject fMRI decoding methods typically train a model on multiple scanned subjects and then adapt it to a new subject using substantial paired fMRI-image data. However, in realistic scenarios, new-subject fMRI data are often limited due to costly data acquisition, and raw data from previous subjects may be inaccessible, leading existing methods to suffer performance degradation during new-subject adaptation. In this paper, we identify that this degradation stems from two key issues: brain-side instability caused by large subject differences in fMRI responses, and image-side supervision unreliability caused by fine-grained visual details that are not reliably supported by limited fMRI signals. To address these challenges, we propose StableMind, a regularized adaptation framework designed to improve brain-side representation stability and image-side supervision reliability. (1) To stabilize brain representations, StableMind reuses ridge projections from the pretrained model as adaptation priors to constrain limited-data new-subject adaptation, and applies Fourier-based feature-level brain augmentation to improve robustness to individual variability. (2) To improve image supervision reliability, StableMind introduces difficulty-aware image blur for brain-image alignment, reducing the influence of fine-grained visual details that are weakly supported by limited fMRI signals while preserving stable visual structure. Experiments on the Natural Scenes Dataset under a unified 1-hour adaptation protocol demonstrate that StableMind achieves 84.02% image retrieval accuracy and 81.66% brain retrieval accuracy averaged over four subjects, surpassing the state-of-the-art method by 5.71% brain retrieval accuracy with fewer trainable adaptation parameters. Our code is available at https://github.com/lingeringlight/StableMind.

preprint2026arXiv

TalkPhoto: A Versatile Training-Free Conversational Assistant for Intelligent Image Editing

Thanks to the powerful language comprehension capabilities of Large Language Models (LLMs), existing instruction-based image editing methods have introduced Multimodal Large Language Models (MLLMs) to promote information exchange between instructions and images, ensuring the controllability and flexibility of image editing. However, these frameworks often build a multi-instruction dataset to train the model to handle multiple editing tasks, which is not only time-consuming and labor-intensive but also fails to achieve satisfactory results. In this paper, we present TalkPhoto, a versatile training-free image editing framework that facilitates precise image manipulation through conversational interaction. We instruct the open-source LLM with a specially designed prompt template to analyze user needs after receiving instructions and hierarchically invoke existing advanced editing methods, all without additional training. Moreover, we implement a plug-and-play and efficient invocation of image editing methods, allowing complex and unseen editing tasks to be integrated into the current framework, achieving stable and high-quality editing results. Extensive experiments demonstrate that our method not only provides more accurate invocation with fewer token consumption but also achieves higher editing quality across various image editing tasks.

preprint2026arXiv

Thinking in Scales: Accelerating Gigapixel Pathology Image Analysis via Adaptive Continuous Reasoning

Traditional whole slide image (WSI) analysis methods typically rely on the multiple instance learning (MIL) paradigm, which extracts patch-level features at high magnification and aggregates them for slide-level prediction. However, such exhaustive patch-level processing is computationally expensive, severely limiting the efficiency and scalability of WSI analysis. To address this challenge, we propose PathCTM (a Pathology-oriented Continuous Thought Model) that enables token-efficient scale-space continuous reasoning for gigapixel WSIs. PathCTM formulates diagnostic inference as a dynamic sequential information pursuit. It progressively transitions from low-magnification global to high-magnification local inspection, and adaptively terminates inference when sufficient evidence is gathered to effectively bound decision uncertainty. Specifically, it uses conditional computation for dynamic scale switching with attention-guided region pruning, coupled with confidence-aware early stopping. Extensive experiments demonstrate that, compared with standard MIL-based methods, PathCTM reduces the number of required image patches by 95.95% and shortens inference time by approximately 95.62%, while maintaining AUC without degradation. Code is available at https://github.com/JSGe-AI/PathCTM.

preprint2026arXiv

Validating Generalist Robots with Situation Calculus and STL Falsification

Generalist robots are becoming a reality, capable of interpreting natural language instructions and executing diverse operations. However, their validation remains challenging because each task induces its own operational context and correctness specification, exceeding the assumptions of traditional validation methods. We propose a two-layer validation framework that combines abstract reasoning with concrete system falsification. At the abstract layer, situation calculus models the world and derives weakest preconditions, enabling constraint-aware combinatorial testing to systematically generate diverse, semantically valid world-task configurations with controllable coverage strength. At the concrete layer, these configurations are instantiated for simulation-based falsification with STL monitoring. Experiments on tabletop manipulation tasks show that our framework effectively uncovers failure cases in the NVIDIA GR00T controller, demonstrating its promise for validating general-purpose robot autonomy.

preprint2026arXiv

Video Generation with Predictive Latents

Video Variational Autoencoder (VAE) enables latent video generative modeling by mapping the visual world into compact spatiotemporal latent spaces, improving training efficiency and stability. While existing video VAEs achieve commendable reconstruction quality, continued optimization of reconstruction does not necessarily translate into improved generative performance. How to enhance the diffusability of video latents remains a critical and unresolved challenge. In this work, inspired by principles of predictive world modeling, we investigate the potential of predictive learning to improve the video generative modeling. To this end, we introduce a simple and effective predictive reconstruction objective that unifies predictive learning with video reconstruction. Specifically, we randomly discard future frames and encode only partial past observations, while training the decoder to reconstruct the observed frames and predict future ones simultaneously. This design encourages the latent space to encode temporally predictive structures and build a more coherent understanding of video dynamics, thereby improving generation quality. Our model, termed Predictive Video VAE (PV-VAE), achieves superior performance on video generation, with 52% faster convergence and a 34.42 FVD improvement over the Wan2.2 VAE on UCF101. Furthermore, comprehensive analyses demonstrate that PV-VAE not only exhibits favorable scalability, with generative performance improving alongside VAE training, but also yields consistent gains in downstream video understanding, underscoring a latent space that effectively captures temporal coherence and motion priors.

preprint2025arXiv

ConstraintLLM: A Neuro-Symbolic Framework for Industrial-Level Constraint Programming

Constraint programming (CP) is a crucial technology for solving real-world constraint optimization problems (COPs), with the advantages of rich modeling semantics and high solving efficiency. Using large language models (LLMs) to generate formal modeling automatically for COPs is becoming a promising approach, which aims to build trustworthy neuro-symbolic AI with the help of symbolic solvers. However, CP has received less attention compared to works based on operations research (OR) models. We introduce ConstraintLLM, the first LLM specifically designed for CP modeling, which is trained on an open-source LLM with multi-instruction supervised fine-tuning. We propose the Constraint-Aware Retrieval Module (CARM) to increase the in-context learning capabilities, which is integrated in a Tree-of-Thoughts (ToT) framework with guided self-correction mechanism. Moreover, we construct and release IndusCP, the first industrial-level benchmark for CP modeling, which contains 140 challenging tasks from various domains. Our experiments demonstrate that ConstraintLLM achieves state-of-the-art solving accuracy across multiple benchmarks and outperforms the baselines by 2x on the new IndusCP benchmark. Code and data are available at: https://github.com/william4s/ConstraintLLM.

preprint2024arXiv

An internet reviews topic hierarchy mining method based on modified continuous renormalization procedure

Mining the hierarchical structure of Internet review topics and realizing a fine classification of review texts can help alleviate users' information overload. However, existing hierarchical topic classification methods primarily rely on external corpora and human intervention. This study proposes a Modified Continuous Renormalization (MCR) procedure that acts on the keyword co-occurrence network with fractal characteristics to achieve the topic hierarchy mining. First, the fractal characteristics in the keyword co-occurrence network of Internet review text are identified using a box-covering algorithm for the first time. Then, the MCR algorithm established on the edge adjacency entropy and the box distance is proposed to obtain the topic hierarchy in the keyword co-occurrence network. Verification data from the Dangdang.com book reviews shows that the MCR constructs topic hierarchies with greater coherence and independence than the HLDA and the Louvain algorithms. Finally, reliable review text classification is achieved using the MCR extended bottom level topic categories. The accuracy rate (P), recall rate (R) and F1 value of Internet review text classification obtained from the MCR-based topic hierarchy are significantly improved compared to four target text classification algorithms.

preprint2024arXiv

Brain-Conditional Multimodal Synthesis: A Survey and Taxonomy

In the era of Artificial Intelligence Generated Content (AIGC), conditional multimodal synthesis technologies (e.g., text-to-image, text-to-video, text-to-audio, etc) are gradually reshaping the natural content in the real world. The key to multimodal synthesis technology is to establish the mapping relationship between different modalities. Brain signals, serving as potential reflections of how the brain interprets external information, exhibit a distinctive One-to-Many correspondence with various external modalities. This correspondence makes brain signals emerge as a promising guiding condition for multimodal content synthesis. Brian-conditional multimodal synthesis refers to decoding brain signals back to perceptual experience, which is crucial for developing practical brain-computer interface systems and unraveling complex mechanisms underlying how the brain perceives and comprehends external stimuli. This survey comprehensively examines the emerging field of AIGC-based Brain-conditional Multimodal Synthesis, termed AIGC-Brain, to delineate the current landscape and future directions. To begin, related brain neuroimaging datasets, functional brain regions, and mainstream generative models are introduced as the foundation of AIGC-Brain decoding and analysis. Next, we provide a comprehensive taxonomy for AIGC-Brain decoding models and present task-specific representative work and detailed implementation strategies to facilitate comparison and in-depth analysis. Quality assessments are then introduced for both qualitative and quantitative evaluation. Finally, this survey explores insights gained, providing current challenges and outlining prospects of AIGC-Brain. Being the inaugural survey in this domain, this paper paves the way for the progress of AIGC-Brain research, offering a foundational overview to guide future work.

preprint2024arXiv

Capacity Results for Multiple-Input Multiple-Output Optical Wireless Communication With Per-Antenna Intensity Constraints

In this paper, we investigate the capacity of a multiple-input multiple-output (MIMO) optical intensity channel (OIC) under per-antenna peak- and average-intensity constraints. We first consider the case where the average intensities of input are required to be equal to preassigned constants due to the requirement of illumination quality and color temperature. When the channel graph of the MIMO OIC is strongly connected, we prove that the strongest eigen-subchannel must have positive channel gains, which simplifies the capacity analysis. Then we derive various capacity bounds by utilizing linear precoding, generalized entropy power inequality, and QR decomposition, etc. These bounds are numerically verified to approach the capacity in the low or high signal-to-noise ratio regime. Specifically, when the channel rank is one less than the number of transmit antennas, we derive an equivalent capacity expression from the perspective of convex geometry, and new lower bounds are derived based on this equivalent expression. Finally, the developed results are extended to the more general case where the average intensities of input are required to be no larger than preassigned constants.

preprint2024arXiv

Hub-collision avoidance and leaf-node options algorithm for fractal dimension and renormalization of complex networks

The box-covering method plays a fundamental role in the fractal property recognition and renormalization analysis of complex networks. This study proposes the hub-collision avoidance and leaf-node options (HALO) algorithm. In the box sampling process, a forward sampling rule (for avoiding hub collisions) and a reverse sampling rule (for preferentially selecting leaf nodes) are determined for bidirectional network traversal to reduce the randomness of sampling. In the box selection process, the larger necessary boxes are preferentially selected to join the solution by continuously removing small boxes. The compact-box-burning (CBB) algorithm, the maximum-excluded-mass-burning (MEMB) algorithm, the overlapping-box-covering (OBCA) algorithm, and the algorithm for combining small-box-removal strategy and maximum box sampling with a sampling density of 30 (SM30) are compared with HALO in experiments. Results on nine real networks show that HALO achieves the highest performance score and obtains 11.40%, 7.67%, 2.18%, and 8.19% fewer boxes than the compared algorithms, respectively. The algorithm determinism is significantly improved. The fractal dimensions estimated by covering four standard networks are more accurate. Moreover, different from MEMB or OBCA, HALO is not affected by the tightness of the hubs and exhibits a stable performance in different networks. Finally, the time complexities of HALO and the compared algorithms are all O(N^2), which is reasonable and acceptable.

preprint2023arXiv

$^{75}$As NMR study of the antiferromagnetic Kondo lattice compound CeNiAsO

We revisit the magnetic properties of the antiferromagnetic Kondo lattice CeNiAsO by $^{75}$As nuclear magnetic resonance measurements. Our results confirm two successive antiferromagnetic transitions of Ce moments at $T_{N1}=9.0(3)$ K and $T_{N2}=7.0(3)$ K. Incommensurate and commensurate antiferromagnetic orders are suggested for $T_{N2}<T<T_{N1}$ and $T<T_{N2}$ respectively, consistent with previous neutron and muon experiments. A Knight shift anomaly, characterized by the failure of $K(T)-χ(T)$ scaling, is observed below $T^*\sim15$ K, which gives a measure of the onset of coherent $c-f$ correlations. This energy scale is further confirmed by the spin-lattice relaxation rate ($1/T_1$). The analysis of spin dynamics also reveals a quasi-two-dimensional character of spin fluctuations in CeNiAsO. This work paves the way for further $^{75}$As nuclear magnetic resonance studies under pressure.

preprint2023arXiv

An Empirical Study on Noisy Label Learning for Program Understanding

Recently, deep learning models have been widely applied in program understanding tasks, and these models achieve state-of-the-art results on many benchmark datasets. A major challenge of deep learning for program understanding is that the effectiveness of these approaches depends on the quality of their datasets, and these datasets often contain noisy data samples. A typical kind of noise in program understanding datasets is label noise, which means that the target outputs for some inputs are incorrect. Researchers have proposed various approaches to alleviate the negative impact of noisy labels, and formed a new research topic: noisy label learning (NLL). In this paper, we conduct an empirical study on the effectiveness of noisy label learning on deep learning for program understanding datasets. We evaluate various NLL approaches and deep learning models on three tasks: program classification, vulnerability detection, and code summarization. From the evaluation results, we come to the following findings: 1) small trained-from-scratch models are prone to label noises in program understanding, while large pre-trained models are highly robust against them. 2) NLL approaches significantly improve the program classification accuracies for small models on noisy training sets, but they only slightly benefit large pre-trained models in classification accuracies. 3) NLL can effectively detect synthetic noises in program understanding, but struggle in detecting real-world noises. We believe our findings can provide insights on the abilities of NLL in program understanding, and shed light on future works in tackling noises in software engineering datasets. We have released our code at https://github.com/jacobwwh/noise_SE.

preprint2023arXiv

Language-Assisted 3D Scene Understanding

The scale and quality of point cloud datasets constrain the advancement of point cloud learning. Recently, with the development of multi-modal learning, the incorporation of domain-agnostic prior knowledge from other modalities, such as images and text, to assist in point cloud feature learning has been considered a promising avenue. Existing methods have demonstrated the effectiveness of multi-modal contrastive training and feature distillation on point clouds. However, challenges remain, including the requirement for paired triplet data, redundancy and ambiguity in supervised features, and the disruption of the original priors. In this paper, we propose a language-assisted approach to point cloud feature learning (LAST-PCL), enriching semantic concepts through LLMs-based text enrichment. We achieve de-redundancy and feature dimensionality reduction without compromising textual priors by statistical-based and training-free significant feature selection. Furthermore, we also delve into an in-depth analysis of the impact of text contrastive training on the point cloud. Extensive experiments validate that the proposed method learns semantically meaningful point cloud features and achieves state-of-the-art or comparable performance in 3D semantic segmentation, 3D object detection, and 3D scene classification tasks.

preprint2023arXiv

Mind Reasoning Manners: Enhancing Type Perception for Generalized Zero-shot Logical Reasoning over Text

Logical reasoning task involves diverse types of complex reasoning over text, based on the form of multiple-choice question answering. Given the context, question and a set of options as the input, previous methods achieve superior performances on the full-data setting. However, the current benchmark dataset has the ideal assumption that the reasoning type distribution on the train split is close to the test split, which is inconsistent with many real application scenarios. To address it, there remain two problems to be studied: (1) How is the zero-shot capability of the models (train on seen types and test on unseen types)? (2) How to enhance the perception of reasoning types for the models? For problem 1, we propose a new benchmark for generalized zero-shot logical reasoning, named ZsLR. It includes six splits based on the three type sampling strategies. For problem 2, a type-aware model TaCo is proposed. It utilizes both the heuristic input reconstruction and the contrastive learning to improve the type perception in the global representation. Extensive experiments on both the zero-shot and full-data settings prove the superiority of TaCo over the state-of-the-art methods. Also, we experiment and verify the generalization capability of TaCo on other logical reasoning dataset.

preprint2022arXiv

A Comprehensive Evaluation of Android ICC Resolution Techniques

Inter-component communication (ICC) is a widely used mechanism in mobile apps, which enables message-based control flow transferring and data passing between Android components. Effective ICC resolution requires precisely identifying entry points, analyzing data values of ICC fields, modeling related framework APIs, etc. Due to various control-flow- and data-flow-related characteristics involved and the lack of oracles for real-world apps, the comprehensive evaluation of ICC resolution techniques is challenging. To fill this gap, we collect multiple-type benchmark suites with 4,104 apps, covering hand-made apps, open-source, and commercial ones. Considering their differences, various evaluation metrics, e.g., number count, graph structure, and reliable oracle based metrics, are adopted on-demand. As the oracle for real-world apps is unavailable, we design a dynamic analysis approach to extract the real ICC links triggered during GUI exploration. By auditing the code implementations, we carefully check the extracted ICCs and confirm 1,680 ones to form a reliable oracle set, in which each ICC is labeled with 25 code characteristic tags. The evaluation performed on six state-of-the-art ICC resolution tools shows that 1) the completeness of static ICC resolution results on real-world apps is not satisfactory, as up to 38%-85% ICCs are missed by tools; 2) many wrongly reported ICCs are sent from or received by only a few components and the graph structure information can help the identification; 3) the efficiency of fundamental tools, like ICC resolution ones, should be optimized in both engineering and research aspects. By investigating both the missed and wrongly reported ICCs, we discuss the strengths of different tools for users and summarize eight common FN/FP patterns in ICC resolution for tool developers.

preprint2022arXiv

A Prompting-based Approach for Adversarial Example Generation and Robustness Enhancement

Recent years have seen the wide application of NLP models in crucial areas such as finance, medical treatment, and news media, raising concerns of the model robustness and vulnerabilities. In this paper, we propose a novel prompt-based adversarial attack to compromise NLP models and robustness enhancement technique. We first construct malicious prompts for each instance and generate adversarial examples via mask-and-filling under the effect of a malicious purpose. Our attack technique targets the inherent vulnerabilities of NLP models, allowing us to generate samples even without interacting with the victim NLP model, as long as it is based on pre-trained language models (PLMs). Furthermore, we design a prompt-based adversarial training method to improve the robustness of PLMs. As our training method does not actually generate adversarial samples, it can be applied to large-scale training sets efficiently. The experimental results show that our attack method can achieve a high attack success rate with more diverse, fluent and natural adversarial examples. In addition, our robustness enhancement method can significantly improve the robustness of models to resist adversarial attacks. Our work indicates that prompting paradigm has great potential in probing some fundamental flaws of PLMs and fine-tuning them for downstream tasks.

preprint2022arXiv

Bayesian Nonparametric Mixtures of Exponential Random Graph Models for Ensembles of Networks

Ensembles of networks arise in various fields where multiple independent networks are observed on the same set of nodes, for example, a collection of brain networks constructed on the same brain regions for different individuals. However, there are few models that describe both the variations and characteristics of networks in an ensemble at the same time. In this paper, we propose to model the ensemble of networks using a Dirichlet Process Mixture of Exponential Random Graph Models (DPM-ERGMs), which divides the ensemble into different clusters and models each cluster of networks using a separate Exponential Random Graph Model (ERGM). By employing a Dirichlet process mixture, the number of clusters can be determined automatically and changed adaptively with the data provided. Moreover, in order to perform full Bayesian inference for DPM-ERGMs, we employ the intermediate importance sampling technique inside the Metropolis-within-slice sampling scheme, which addressed the problem of sampling from the intractable ERGMs on an infinite sample space. We also demonstrate the performance of DPM-ERGMs with both simulated and real datasets.

preprint2022arXiv

Conceptual Compression via Deep Structure and Texture Synthesis

Existing compression methods typically focus on the removal of signal-level redundancies, while the potential and versatility of decomposing visual data into compact conceptual components still lack further study. To this end, we propose a novel conceptual compression framework that encodes visual data into compact structure and texture representations, then decodes in a deep synthesis fashion, aiming to achieve better visual reconstruction quality, flexible content manipulation, and potential support for various vision tasks. In particular, we propose to compress images by a dual-layered model consisting of two complementary visual features: 1) structure layer represented by structural maps and 2) texture layer characterized by low-dimensional deep representations. At the encoder side, the structural maps and texture representations are individually extracted and compressed, generating the compact, interpretable, inter-operable bitstreams. During the decoding stage, a hierarchical fusion GAN (HF-GAN) is proposed to learn the synthesis paradigm where the textures are rendered into the decoded structural maps, leading to high-quality reconstruction with remarkable visual realism. Extensive experiments on diverse images have demonstrated the superiority of our framework with lower bitrates, higher reconstruction quality, and increased versatility towards visual analysis and content manipulation tasks.

preprint2022arXiv

Content-aware Scalable Deep Compressed Sensing

To more efficiently address image compressed sensing (CS) problems, we present a novel content-aware scalable network dubbed CASNet which collectively achieves adaptive sampling rate allocation, fine granular scalability and high-quality reconstruction. We first adopt a data-driven saliency detector to evaluate the importances of different image regions and propose a saliency-based block ratio aggregation (BRA) strategy for sampling rate allocation. A unified learnable generating matrix is then developed to produce sampling matrix of any CS ratio with an ordered structure. Being equipped with the optimization-inspired recovery subnet guided by saliency information and a multi-block training scheme preventing blocking artifacts, CASNet jointly reconstructs the image blocks sampled at various sampling rates with one single model. To accelerate training convergence and improve network robustness, we propose an SVD-based initialization scheme and a random transformation enhancement (RTE) strategy, which are extensible without introducing extra parameters. All the CASNet components can be combined and learned end-to-end. We further provide a four-stage implementation for evaluation and practical deployments. Experiments demonstrate that CASNet outperforms other CS networks by a large margin, validating the collaboration and mutual supports among its components and strategies. Codes are available at https://github.com/Guaishou74851/CASNet.

preprint2022arXiv

Deep Generalized Unfolding Networks for Image Restoration

Deep neural networks (DNN) have achieved great success in image restoration. However, most DNN methods are designed as a black box, lacking transparency and interpretability. Although some methods are proposed to combine traditional optimization algorithms with DNN, they usually demand pre-defined degradation processes or handcrafted assumptions, making it difficult to deal with complex and real-world applications. In this paper, we propose a Deep Generalized Unfolding Network (DGUNet) for image restoration. Concretely, without loss of interpretability, we integrate a gradient estimation strategy into the gradient descent step of the Proximal Gradient Descent (PGD) algorithm, driving it to deal with complex and real-world image degradation. In addition, we design inter-stage information pathways across proximal mapping in different PGD iterations to rectify the intrinsic information loss in most deep unfolding networks (DUN) through a multi-scale and spatial-adaptive way. By integrating the flexible gradient descent and informative proximal mapping, we unfold the iterative PGD algorithm into a trainable DNN. Extensive experiments on various image restoration tasks demonstrate the superiority of our method in terms of state-of-the-art performance, interpretability, and generalizability. The source code is available at https://github.com/MC-E/Deep-Generalized-Unfolding-Networks-for-Image-Restoration.

preprint2022arXiv

Elastic properties of a Sc-Zr-Nb-Ta-Rh-Pd high-entropy alloy superconductor

We report a comprehensive study on the elastic properties of a hexanary high-entropy alloy superconductor (ScZrNbTa)$_{0.685}$[RhPd]$_{0.315}$ at room and cryogenic temperatures, by Resonant Ultrasound Spectroscopy experiments. The derived elastic constants are bulk modulus $K=132.7$ GPa, Young's modulus $E=121.0$ GPa, shear modulus $G=44.9$ GPa, and Poisson's ratio $ν$=0.348 for room temperature. The Young's and shear moduli are $\sim 10\%$ larger than those in NbTi superconductor with similar $T_c$, while the ductility is comparable. Moreover, the mechanical performance is further enhanced at cryogenic temperature. Our work confirms the advantageous mechanical properties of high-entropy alloy superconductors and suggests the application prospects.

preprint2022arXiv

Exploring fundamental laws of classical mechanics via predicting the orbits of planets based on neural networks

Neural networks have provided powerful approaches to solve various scientific problems. Many of them are even difficult for human experts who are good at accessing the physical laws from experimental data. We investigate whether neural networks can assist us in exploring the fundamental laws of classical mechanics from data of planetary motion. Firstly, we predict the orbits of planets in the geocentric system using the gate recurrent unit, one of the common neural networks. We find that the precision of the prediction is obviously improved when the information of the Sun is included in the training set. This result implies that the Sun is particularly important in the geocentric system without any prior knowledge, which inspires us to gain Copernicus' heliocentric theory. Secondly, we turn to the heliocentric system and make successfully mutual predictions between the position and velocity of planets. We hold that the successful prediction is due to the existence of enough conserved quantities (such as conservations of mechanical energy and angular momentum) in the system. Our research provides a new way to explore the existence of conserved quantities in mechanics system based on neural networks.

preprint2022arXiv

Frequency Domain Model Augmentation for Adversarial Attack

For black-box attacks, the gap between the substitute model and the victim model is usually large, which manifests as a weak attack performance. Motivated by the observation that the transferability of adversarial examples can be improved by attacking diverse models simultaneously, model augmentation methods which simulate different models by using transformed images are proposed. However, existing transformations for spatial domain do not translate to significantly diverse augmented models. To tackle this issue, we propose a novel spectrum simulation attack to craft more transferable adversarial examples against both normally trained and defense models. Specifically, we apply a spectrum transformation to the input and thus perform the model augmentation in the frequency domain. We theoretically prove that the transformation derived from frequency domain leads to a diverse spectrum saliency map, an indicator we proposed to reflect the diversity of substitute models. Notably, our method can be generally combined with existing attacks. Extensive experiments on the ImageNet dataset demonstrate the effectiveness of our method, \textit{e.g.}, attacking nine state-of-the-art defense models with an average success rate of \textbf{95.4\%}. Our code is available in \url{https://github.com/yuyang-long/SSA}.

preprint2022arXiv

HerosNet: Hyperspectral Explicable Reconstruction and Optimal Sampling Deep Network for Snapshot Compressive Imaging

Hyperspectral imaging is an essential imaging modality for a wide range of applications, especially in remote sensing, agriculture, and medicine. Inspired by existing hyperspectral cameras that are either slow, expensive, or bulky, reconstructing hyperspectral images (HSIs) from a low-budget snapshot measurement has drawn wide attention. By mapping a truncated numerical optimization algorithm into a network with a fixed number of phases, recent deep unfolding networks (DUNs) for spectral snapshot compressive sensing (SCI) have achieved remarkable success. However, DUNs are far from reaching the scope of industrial applications limited by the lack of cross-phase feature interaction and adaptive parameter adjustment. In this paper, we propose a novel Hyperspectral Explicable Reconstruction and Optimal Sampling deep Network for SCI, dubbed HerosNet, which includes several phases under the ISTA-unfolding framework. Each phase can flexibly simulate the sensing matrix and contextually adjust the step size in the gradient descent step, and hierarchically fuse and interact the hidden states of previous phases to effectively recover current HSI frames in the proximal mapping step. Simultaneously, a hardware-friendly optimal binary mask is learned end-to-end to further improve the reconstruction performance. Finally, our HerosNet is validated to outperform the state-of-the-art methods on both simulation and real datasets by large margins. The source code is available at https://github.com/jianzhangcs/HerosNet.

preprint2022arXiv

Horizontal and Vertical Attention in Transformers

Transformers are built upon multi-head scaled dot-product attention and positional encoding, which aim to learn the feature representations and token dependencies. In this work, we focus on enhancing the distinctive representation by learning to augment the feature maps with the self-attention mechanism in Transformers. Specifically, we propose the horizontal attention to re-weight the multi-head output of the scaled dot-product attention before dimensionality reduction, and propose the vertical attention to adaptively re-calibrate channel-wise feature responses by explicitly modelling inter-dependencies among different channels. We demonstrate the Transformer models equipped with the two attentions have a high generalization capability across different supervised learning tasks, with a very minor additional computational cost overhead. The proposed horizontal and vertical attentions are highly modular, which can be inserted into various Transformer models to further improve the performance. Our code is available in the supplementary material.

preprint2022arXiv

Improving Tese Case Generation for Python Native Libraries Through Constraints on Input Data Structures

Modern Python projects execute computational functions using native libraries and give Python interfaces to boost execution speed; hence, testing these libraries becomes critical to the project's robustness. One challenge is that existing approaches use coverage to guide generation, but native libraries run as black boxes to Python code with no execution information. Another is that dynamic binary instrumentation reduces testing performance as it needs to monitor both native libraries and the Python virtual machine. To address these challenges, in this paper, we propose an automated test case generation approach that works at the Python code layer. Our insight is that many path conditions in native libraries are for processing input data structures through interacting with the VM. In our approach, we instrument the Python Interpreter to monitor the interactions between native libraries and VM, derive constraints on the structures, and then use the constraints to guide test case generation. We implement our approach in a tool named PyCing and apply it to six widely-used Python projects. The experimental results reveal that with the structure constraint guidance, PyCing can cover more execution paths than existing test cases and state-of-the-art tools. Also, with the checkers in the testing framework Pytest, PyCing can identify segmentation faults in 10 Python interfaces and memory leaks in 9. Our instrumentation strategy also has an acceptable influence on testing efficiency.

preprint2022arXiv

Intrinsic new properties of a quantum spin liquid

Quantum fluctuations are expected to lead to highly entangled spin-liquid states in certain two-dimensional spin-1/2 compounds. We have synthesized and measured thermodynamic properties and muon spin relaxation rates in the copper-based two-dimensional triangular-lattice spin liquids Lu$_3$Cu$_2$Sb$_3$O$_{14}$ and Lu$_3$CuZnSb$_3$O$_{14}$. The former is the least disordered of this kind discovered to date. Magnetic entropy generation at high temperatures has been ruled out after carefully correcting for the lattice specific heat. Surprisingly, roughly half of the magnetic entropy is missing down to temperatures of O(10$^{-3}$) the exchange energy, independent of magnetic field up to $gμ_B H \gtrsim k_BΘ_W$, where $Θ_W$ is the Weiss temperature. The magnetic specific heat divided by temperature $C_M(T)/T$ and muon spin relaxation rate $λ(T)$ are both temperature-independent at low temperatures, followed by logarithmic decreases with increasing temperature. This behavior can be simply characterized by scale-invariant time-dependent fluctuations with a single parameter. Since no cooperative effects due to impurities are observed, the measured properties are intrinsic. They are evidence that in Lu$_3$Cu$_2$Sb$_3$O$_{14}$ massive quantum fluctuations lead to either a gigantic specific heat peak from singlet excitations at very low temperatures or, perhaps less likely, an extensively degenerate possibly topological singlet ground state.

preprint2022arXiv

Investigating Accuracy-Novelty Performance for Graph-based Collaborative Filtering

Recent years have witnessed the great accuracy performance of graph-based Collaborative Filtering (CF) models for recommender systems. By taking the user-item interaction behavior as a graph, these graph-based CF models borrow the success of Graph Neural Networks (GNN), and iteratively perform neighborhood aggregation to propagate the collaborative signals. While conventional CF models are known for facing the challenges of the popularity bias that favors popular items, one may wonder "Whether the existing graph-based CF models alleviate or exacerbate popularity bias of recommender systems?" To answer this question, we first investigate the two-fold performances w.r.t. accuracy and novelty for existing graph-based CF methods. The empirical results show that symmetric neighborhood aggregation adopted by most existing graph-based CF models exacerbate the popularity bias and this phenomenon becomes more serious as the depth of graph propagation increases. Further, we theoretically analyze the cause of popularity bias for graph-based CF. Then, we propose a simple yet effective plugin, namely r-AdjNorm, to achieve an accuracy-novelty trade-off by controlling the normalization strength in the neighborhood aggregation process. Meanwhile, r-AdjNorm can be smoothly applied to the existing graph-based CF backbones without additional computation. Finally, experimental results on three benchmark datasets show that our proposed method can improve novelty without sacrificing accuracy under various graph-based CF backbones.

preprint2022arXiv

Low-Memory Neural Network Training: A Technical Report

Memory is increasingly often the bottleneck when training neural network models. Despite this, techniques to lower the overall memory requirements of training have been less widely studied compared to the extensive literature on reducing the memory requirements of inference. In this paper we study a fundamental question: How much memory is actually needed to train a neural network? To answer this question, we profile the overall memory usage of training on two representative deep learning benchmarks -- the WideResNet model for image classification and the DynamicConv Transformer model for machine translation -- and comprehensively evaluate four standard techniques for reducing the training memory requirements: (1) imposing sparsity on the model, (2) using low precision, (3) microbatching, and (4) gradient checkpointing. We explore how each of these techniques in isolation affects both the peak memory usage of training and the quality of the end model, and explore the memory, accuracy, and computation tradeoffs incurred when combining these techniques. Using appropriate combinations of these techniques, we show that it is possible to the reduce the memory required to train a WideResNet-28-2 on CIFAR-10 by up to 60.7x with a 0.4% loss in accuracy, and reduce the memory required to train a DynamicConv model on IWSLT'14 German to English translation by up to 8.7x with a BLEU score drop of 0.15.

preprint2022arXiv

Measuring and Improving the Use of Graph Information in Graph Neural Networks

Graph neural networks (GNNs) have been widely used for representation learning on graph data. However, there is limited understanding on how much performance GNNs actually gain from graph data. This paper introduces a context-surrounding GNN framework and proposes two smoothness metrics to measure the quantity and quality of information obtained from graph data. A new GNN model, called CS-GNN, is then designed to improve the use of graph information based on the smoothness values of a graph. CS-GNN is shown to achieve better performance than existing methods in different types of real graphs.

preprint2022arXiv

MM-RealSR: Metric Learning based Interactive Modulation for Real-World Super-Resolution

Interactive image restoration aims to restore images by adjusting several controlling coefficients, which determine the restoration strength. Existing methods are restricted in learning the controllable functions under the supervision of known degradation types and levels. They usually suffer from a severe performance drop when the real degradation is different from their assumptions. Such a limitation is due to the complexity of real-world degradations, which can not provide explicit supervision to the interactive modulation during training. However, how to realize the interactive modulation in real-world super-resolution has not yet been studied. In this work, we present a Metric Learning based Interactive Modulation for Real-World Super-Resolution (MM-RealSR). Specifically, we propose an unsupervised degradation estimation strategy to estimate the degradation level in real-world scenarios. Instead of using known degradation levels as explicit supervision to the interactive mechanism, we propose a metric learning strategy to map the unquantifiable degradation levels in real-world scenarios to a metric space, which is trained in an unsupervised manner. Moreover, we introduce an anchor point strategy in the metric learning process to normalize the distribution of metric space. Extensive experiments demonstrate that the proposed MM-RealSR achieves excellent modulation and restoration performance in real-world super-resolution. Codes are available at https://github.com/TencentARC/MM-RealSR.

preprint2022arXiv

Multiple Instance Learning with Mixed Supervision in Gleason Grading

With the development of computational pathology, deep learning methods for Gleason grading through whole slide images (WSIs) have excellent prospects. Since the size of WSIs is extremely large, the image label usually contains only slide-level label or limited pixel-level labels. The current mainstream approach adopts multi-instance learning to predict Gleason grades. However, some methods only considering the slide-level label ignore the limited pixel-level labels containing rich local information. Furthermore, the method of additionally considering the pixel-level labels ignores the inaccuracy of pixel-level labels. To address these problems, we propose a mixed supervision Transformer based on the multiple instance learning framework. The model utilizes both slide-level label and instance-level labels to achieve more accurate Gleason grading at the slide level. The impact of inaccurate instance-level labels is further reduced by introducing an efficient random masking strategy in the mixed supervision training process. We achieve the state-of-the-art performance on the SICAPv2 dataset, and the visual analysis shows the accurate prediction results of instance level. The source code is available at https://github.com/bianhao123/Mixed_supervision.

preprint2022arXiv

MVDG: A Unified Multi-view Framework for Domain Generalization

To generalize the model trained in source domains to unseen target domains, domain generalization (DG) has recently attracted lots of attention. Since target domains can not be involved in training, overfitting source domains is inevitable. As a popular regularization technique, the meta-learning training scheme has shown its ability to resist overfitting. However, in the training stage, current meta-learning-based methods utilize only one task along a single optimization trajectory, which might produce a biased and noisy optimization direction. Beyond the training stage, overfitting could also cause unstable prediction in the test stage. In this paper, we propose a novel multi-view DG framework to effectively reduce the overfitting in both the training and test stage. Specifically, in the training stage, we develop a multi-view regularized meta-learning algorithm that employs multiple optimization trajectories to produce a suitable optimization direction for model updating. We also theoretically show that the generalization bound could be reduced by increasing the number of tasks in each trajectory. In the test stage, we utilize multiple augmented images to yield a multi-view prediction to alleviate unstable prediction, which significantly promotes model reliability. Extensive experiments on three benchmark datasets validate that our method can find a flat minimum to enhance generalization and outperform several state-of-the-art approaches.

preprint2022arXiv

NeRFocus: Neural Radiance Field for 3D Synthetic Defocus

Neural radiance fields (NeRF) bring a new wave for 3D interactive experiences. However, as an important part of the immersive experiences, the defocus effects have not been fully explored within NeRF. Some recent NeRF-based methods generate 3D defocus effects in a post-process fashion by utilizing multiplane technology. Still, they are either time-consuming or memory-consuming. This paper proposes a novel thin-lens-imaging-based NeRF framework that can directly render various 3D defocus effects, dubbed NeRFocus. Unlike the pinhole, the thin lens refracts rays of a scene point, so its imaging on the sensor plane is scattered as a circle of confusion (CoC). A direct solution sampling enough rays to approximate this process is computationally expensive. Instead, we propose to inverse the thin lens imaging to explicitly model the beam path for each point on the sensor plane and generalize this paradigm to the beam path of each pixel, then use the frustum-based volume rendering to render each pixel's beam path. We further design an efficient probabilistic training (p-training) strategy to simplify the training process vastly. Extensive experiments demonstrate that our NeRFocus can achieve various 3D defocus effects with adjustable camera pose, focus distance, and aperture size. Existing NeRF can be regarded as our special case by setting aperture size as zero to render large depth-of-field images. Despite such merits, NeRFocus does not sacrifice NeRF's original performance (e.g., training and inference time, parameter consumption, rendering quality), which implies its great potential for broader application and further improvement. Code and video are available at https://github.com/wyhuai/NeRFocus.

preprint2022arXiv

Panini-Net: GAN Prior Based Degradation-Aware Feature Interpolation for Face Restoration

Emerging high-quality face restoration (FR) methods often utilize pre-trained GAN models (\textit{i.e.}, StyleGAN2) as GAN Prior. However, these methods usually struggle to balance realness and fidelity when facing various degradation levels. Besides, there is still a noticeable visual quality gap compared with pre-trained GAN models. In this paper, we propose a novel GAN Prior based degradation-aware feature interpolation network, dubbed Panini-Net, for FR tasks by explicitly learning the abstract representations to distinguish various degradations. Specifically, an unsupervised degradation representation learning (UDRL) strategy is first developed to extract degradation representations (DR) of the input degraded images. Then, a degradation-aware feature interpolation (DAFI) module is proposed to dynamically fuse the two types of informative features (\textit{i.e.}, features from input images and features from GAN Prior) with flexible adaption to various degradations based on DR. Ablation studies reveal the working mechanism of DAFI and its potential for editable FR. Extensive experiments demonstrate that our Panini-Net achieves state-of-the-art performance for multi-degradation face restoration and face super-resolution. The source code is available at https://github.com/jianzhangcs/panini.

preprint2022arXiv

PUERT: Probabilistic Under-sampling and Explicable Reconstruction Network for CS-MRI

Compressed Sensing MRI (CS-MRI) aims at reconstructing de-aliased images from sub-Nyquist sampling k-space data to accelerate MR Imaging, thus presenting two basic issues, i.e., where to sample and how to reconstruct. To deal with both problems simultaneously, we propose a novel end-to-end Probabilistic Under-sampling and Explicable Reconstruction neTwork, dubbed PUERT, to jointly optimize the sampling pattern and the reconstruction network. Instead of learning a deterministic mask, the proposed sampling subnet explores an optimal probabilistic sub-sampling pattern, which describes independent Bernoulli random variables at each possible sampling point, thus retaining robustness and stochastics for a more reliable CS reconstruction. A dynamic gradient estimation strategy is further introduced to gradually approximate the binarization function in backward propagation, which efficiently preserves the gradient information and further improves the reconstruction quality. Moreover, in our reconstruction subnet, we adopt a model-based network design scheme with high efficiency and interpretability, which is shown to assist in further exploitation for the sampling subnet. Extensive experiments on two widely used MRI datasets demonstrate that our proposed PUERT not only achieves state-of-the-art results in terms of both quantitative metrics and visual quality but also yields a sub-sampling pattern and a reconstruction model that are both customized to training data.

preprint2022arXiv

Quantifying Robustness to Adversarial Word Substitutions

Deep-learning-based NLP models are found to be vulnerable to word substitution perturbations. Before they are widely adopted, the fundamental issues of robustness need to be addressed. Along this line, we propose a formal framework to evaluate word-level robustness. First, to study safe regions for a model, we introduce robustness radius which is the boundary where the model can resist any perturbation. As calculating the maximum robustness radius is computationally hard, we estimate its upper and lower bound. We repurpose attack methods as ways of seeking upper bound and design a pseudo-dynamic programming algorithm for a tighter upper bound. Then verification method is utilized for a lower bound. Further, for evaluating the robustness of regions outside a safe radius, we reexamine robustness from another view: quantification. A robustness metric with a rigorous statistical guarantee is introduced to measure the quantification of adversarial examples, which indicates the model's susceptibility to perturbations outside the safe radius. The metric helps us figure out why state-of-the-art models like BERT can be easily fooled by a few word substitutions, but generalize well in the presence of real-world noises.

preprint2022arXiv

R-DFCIL: Relation-Guided Representation Learning for Data-Free Class Incremental Learning

Class-Incremental Learning (CIL) struggles with catastrophic forgetting when learning new knowledge, and Data-Free CIL (DFCIL) is even more challenging without access to the training data of previously learned classes. Though recent DFCIL works introduce techniques such as model inversion to synthesize data for previous classes, they fail to overcome forgetting due to the severe domain gap between the synthetic and real data. To address this issue, this paper proposes relation-guided representation learning (RRL) for DFCIL, dubbed R-DFCIL. In RRL, we introduce relational knowledge distillation to flexibly transfer the structural relation of new data from the old model to the current model. Our RRL-boosted DFCIL can guide the current model to learn representations of new classes better compatible with representations of previous classes, which greatly reduces forgetting while improving plasticity. To avoid the mutual interference between representation and classifier learning, we employ local rather than global classification loss during RRL. After RRL, the classification head is refined with global class-balanced classification loss to address the data imbalance issue as well as learn the decision boundaries between new and previous classes. Extensive experiments on CIFAR100, Tiny-ImageNet200, and ImageNet100 demonstrate that our R-DFCIL significantly surpasses previous approaches and achieves a new state-of-the-art performance for DFCIL. Code is available at https://github.com/jianzhangcs/R-DFCIL

preprint2022arXiv

Ray Priors through Reprojection: Improving Neural Radiance Fields for Novel View Extrapolation

Neural Radiance Fields (NeRF) have emerged as a potent paradigm for representing scenes and synthesizing photo-realistic images. A main limitation of conventional NeRFs is that they often fail to produce high-quality renderings under novel viewpoints that are significantly different from the training viewpoints. In this paper, instead of exploiting few-shot image synthesis, we study the novel view extrapolation setting that (1) the training images can well describe an object, and (2) there is a notable discrepancy between the training and test viewpoints' distributions. We present RapNeRF (RAy Priors) as a solution. Our insight is that the inherent appearances of a 3D surface's arbitrary visible projections should be consistent. We thus propose a random ray casting policy that allows training unseen views using seen views. Furthermore, we show that a ray atlas pre-computed from the observed rays' viewing directions could further enhance the rendering quality for extrapolated views. A main limitation is that RapNeRF would remove the strong view-dependent effects because it leverages the multi-view consistency property.

preprint2022arXiv

Research on the Inverse Kinematics Prediction of a Soft Biomimetic Actuator via BP Neural Network

In this work, we address the inverse kinetics problem of motion planning of soft biomimetic actuators driven by three chambers. Soft biomimetic actuators have been applied in many applications owing to their intrinsic softness. Although a mathematical model can be derived to describe the inverse dynamics of this actuator, it is still not accurate to capture the nonlinearity and uncertainty of the material and the system. Besides, such a complex model is time-consuming, so it is not easy to apply in the real-time control unit. Therefore, developing a model-free approach in this area could be a new idea. To overcome these intrinsic problems, we propose a back-propagation (BP) neural network learning the inverse kinetics of the soft biomimetic actuator moving in three-dimensional space. After training with sample data, the BP neural network model can represent the relation between the manipulator tip position and the pressure applied to the chambers. The proposed algorithm is more precise than the analytical model. The results show that a desired terminal position can be achieved with a degree of accuracy of 2.46% relative average error with respect to the total actuator length.

preprint2022arXiv

Self-attention on Multi-Shifted Windows for Scene Segmentation

Scene segmentation in images is a fundamental yet challenging problem in visual content understanding, which is to learn a model to assign every image pixel to a categorical label. One of the challenges for this learning task is to consider the spatial and semantic relationships to obtain descriptive feature representations, so learning the feature maps from multiple scales is a common practice in scene segmentation. In this paper, we explore the effective use of self-attention within multi-scale image windows to learn descriptive visual features, then propose three different strategies to aggregate these feature maps to decode the feature representation for dense prediction. Our design is based on the recently proposed Swin Transformer models, which totally discards convolution operations. With the simple yet effective multi-scale feature learning and aggregation, our models achieve very promising performance on four public scene segmentation datasets, PASCAL VOC2012, COCO-Stuff 10K, ADE20K and Cityscapes.

preprint2022arXiv

Series Photo Selection via Multi-view Graph Learning

Series photo selection (SPS) is an important branch of the image aesthetics quality assessment, which focuses on finding the best one from a series of nearly identical photos. While a great progress has been observed, most of the existing SPS approaches concentrate solely on extracting features from the original image, neglecting that multiple views, e.g, saturation level, color histogram and depth of field of the image, will be of benefit to successfully reflecting the subtle aesthetic changes. Taken multi-view into consideration, we leverage a graph neural network to construct the relationships between multi-view features. Besides, multiple views are aggregated with an adaptive-weight self-attention module to verify the significance of each view. Finally, a siamese network is proposed to select the best one from a series of nearly identical photos. Experimental results demonstrate that our model accomplish the highest success rates compared with competitive methods.

preprint2022arXiv

Small Solitons and Multi-Solitons in Generalized Davey-Stewartson System

This paper is concerned with the generalized Davey-Stewarston system in two dimensional space. Existence and stability of small solitons are proved by solving two correlative constrained variational problems and spectrum analysis. In addition, multi-solitons with different speeds are constructed by bootstrap argument.

preprint2022arXiv

Solving Large-Scale Dynamic Vehicle Routing Problems with Stochastic Requests

Dynamic vehicle routing problems (DVRPs) arise in several applications such as technician routing, meal delivery, and parcel shipping. We consider the DVRP with stochastic customer requests (DVRPSR), in which vehicles must be routed dynamically with the goal of maximizing the number of served requests. We model the DVRPSR as a multi-stage optimization problem, where the first-stage decision defines route plans for serving scheduled requests. Our main contributions are knapsack-based linear models to approximate accurately the expected reward-to-go, measured as the number of accepted requests, at any state of the stochastic system. These approximations are based on representing each vehicle as a knapsack with a capacity given by the remaining service time available along the vehicle's route. We combine these approximations with optimal acceptance and assignment decision rules and derive efficient and high-performing online scheduling policies. We further leverage good predictions of the expected reward-to-go to design initial route plans that facilitate serving dynamic requests. Computational experiments on very large instances based on a real street network demonstrate the effectiveness of the proposed methods in prescribing high-quality offline route plans and online scheduling decisions.

preprint2022arXiv

Stability of Big Solitons in a Competitive Power Nonlinear Schrödinger Equation

By introducing and solving two correlative constrained variational problems as well as spectrum analysis, an approach to fix soliton frequency from the prescribed mass for nonlinear Schrödinger equations is found, and an open problem in normalized solutions is answered. Then existence and orbital stability of big solitons depending on frequencies for nonlinear Schrödinger equation with competitive power nonlinearity is proved for the first time. In addition multi-solitons of the equation with different speeds are constructed by stable big solitons.

preprint2022arXiv

The Foreseeable Future: Self-Supervised Learning to Predict Dynamic Scenes for Indoor Navigation

We present a method for generating, predicting, and using Spatiotemporal Occupancy Grid Maps (SOGM), which embed future semantic information of real dynamic scenes. We present an auto-labeling process that creates SOGMs from noisy real navigation data. We use a 3D-2D feedforward architecture, trained to predict the future time steps of SOGMs, given 3D lidar frames as input. Our pipeline is entirely self-supervised, thus enabling lifelong learning for real robots. The network is composed of a 3D back-end that extracts rich features and enables the semantic segmentation of the lidar frames, and a 2D front-end that predicts the future information embedded in the SOGM representation, potentially capturing the complexities and uncertainties of real-world multi-agent, multi-future interactions. We also design a navigation system that uses these predicted SOGMs within planning, after they have been transformed into Spatiotemporal Risk Maps (SRMs). We verify our navigation system's abilities in simulation, validate it on a real robot, study SOGM predictions on real data in various circumstances, and provide a novel indoor 3D lidar dataset, collected during our experiments, which includes our automated annotations.

preprint2022arXiv

TransCL: Transformer Makes Strong and Flexible Compressive Learning

Compressive learning (CL) is an emerging framework that integrates signal acquisition via compressed sensing (CS) and machine learning for inference tasks directly on a small number of measurements. It can be a promising alternative to classical image-domain methods and enjoys great advantages in memory saving and computational efficiency. However, previous attempts on CL are not only limited to a fixed CS ratio, which lacks flexibility, but also limited to MNIST/CIFAR-like datasets and do not scale to complex real-world high-resolution (HR) data or vision tasks. In this paper, a novel transformer-based compressive learning framework on large-scale images with arbitrary CS ratios, dubbed TransCL, is proposed. Specifically, TransCL first utilizes the strategy of learnable block-based compressed sensing and proposes a flexible linear projection strategy to enable CL to be performed on large-scale images in an efficient block-by-block manner with arbitrary CS ratios. Then, regarding CS measurements from all blocks as a sequence, a pure transformer-based backbone is deployed to perform vision tasks with various task-oriented heads. Our sufficient analysis presents that TransCL exhibits strong resistance to interference and robust adaptability to arbitrary CS ratios. Extensive experiments for complex HR data demonstrate that the proposed TransCL can achieve state-of-the-art performance in image classification and semantic segmentation tasks. In particular, TransCL with a CS ratio of $10\%$ can obtain almost the same performance as when operating directly on the original data and can still obtain satisfying performance even with an extremely low CS ratio of $1\%$. The source codes of our proposed TransCL is available at \url{https://github.com/MC-E/TransCL/}.

preprint2022arXiv

Ultimately-scaled electrodes for contacting individual atomically-precise graphene nanoribbons

Bottom-up synthesized graphene nanoribbons (GNRs) are quantum materials that can be structured with atomic precision, providing unprecedented control over their physical properties. Accessing the intrinsic functionality of GNRs for quantum technology applications requires the manipulation of single charges, spins, or photons at the level of an individual GNR. However, experimentally, contacting individual GNRs remains challenging due to their nanometer-sized width and length as well as their high density on the metallic growth substrate. Here, we demonstrate the contacting and electrical characterization of individual GNRs in a multi-gate device architecture using single-walled carbon nanotubes (SWNTs) as ultimately-scaled electrodes. The GNR-SWNT devices exhibit well-defined quantum transport phenomena, including Coulomb blockade, excited states, and Franck-Condon blockade, all characteristics pointing towards the contacting of an individual GNR. Combined with the multi-gate architecture, this contacting method opens a road for the integration of GNRs in quantum devices to exploit their topologically trivial and non-trivial nature.

preprint2022arXiv

Uniaxial stress effect on the quasi-one-dimensional Kondo lattice CeCo$_2$Ga$_8$

Quantum critical phenomena in the quasi-one-dimensional limit remains an open issue. We report the uniaxial stress effect on the quasi-one-dimensional Kondo lattice CeCo$_2$Ga$_8$ by electric transport and AC heat capacity measurements. CeCo$_2$Ga$_8$ is speculated to sit in close vicinity but on the quantum-disordered side of a quantum critical point. Upon compressing the ${c}$ axis, parallel to the Ce-Ce chain, the onset of coherent Kondo effect is enhanced. In contrast, the electronic specific heat diverges more rapidly at low temperature when the intra-chain distance is elongated by compressions along {a} or {b} axes. These results suggest that a tensile intra-chain strain ($\varepsilon_c >0$) pushes CeCo$_2$Ga$_8$ closer to a quantum critical point, while a compressive intra-chain strain ($\varepsilon_c <0$) likely causes departure. Our work provides a rare paradigm of manipulation near a quantum critical point in a quasi-1D Kondo lattice by uniaxial stress, and paves the way for further investigations on the unique feature of quantum criticality in the quasi-1D limit.

preprint2022arXiv

Unsupervised Part Mining for Fine-grained Image Classification

Fine-grained image classification remains challenging due to the large intra-class variance and small inter-class variance. Since the subtle visual differences are only in local regions of discriminative parts among subcategories, part localization is a key issue for fine-grained image classification. Most existing approaches localize object or parts in an image with object or part annotations, which are expensive and labor-consuming. To tackle this issue, we propose a fully unsupervised part mining (UPM) approach to localize the discriminative parts without even image-level annotations, which largely improves the fine-grained classification performance. We first utilize pattern mining techniques to discover frequent patterns, i.e., co-occurrence highlighted regions, in the feature maps extracted from a pre-trained convolutional neural network (CNN) model. Inspired by the fact that these relevant meaningful patterns typically hold appearance and spatial consistency, we then cluster the mined regions to obtain the cluster centers and the discriminative parts surrounding the cluster centers are generated. Importantly, any annotations and sophisticated training procedures are not used in our proposed part localization approach. Finally, a multi-stream classification network is built for aggregating the original, object-level and part-level features simultaneously. Compared with other state-of-the-art approaches, our UPM approach achieves the competitive performance.

preprint2022arXiv

Using EBGAN for Anomaly Intrusion Detection

As an active network security protection scheme, intrusion detection system (IDS) undertakes the important responsibility of detecting network attacks in the form of malicious network traffic. Intrusion detection technology is an important part of IDS. At present, many scholars have carried out extensive research on intrusion detection technology. However, developing an efficient intrusion detection method for massive network traffic data is still difficult. Since Generative Adversarial Networks (GANs) have powerful modeling capabilities for complex high-dimensional data, they provide new ideas for addressing this problem. In this paper, we put forward an EBGAN-based intrusion detection method, IDS-EBGAN, that classifies network records as normal traffic or malicious traffic. The generator in IDS-EBGAN is responsible for converting the original malicious network traffic in the training set into adversarial malicious examples. This is because we want to use adversarial learning to improve the ability of discriminator to detect malicious traffic. At the same time, the discriminator adopts Autoencoder model. During testing, IDS-EBGAN uses reconstruction error of discriminator to classify traffic records.

preprint2021arXiv

A comprehensive survey on point cloud registration

Registration is a transformation estimation problem between two point clouds, which has a unique and critical role in numerous computer vision applications. The developments of optimization-based methods and deep learning methods have improved registration robustness and efficiency. Recently, the combinations of optimization-based and deep learning methods have further improved performance. However, the connections between optimization-based and deep learning methods are still unclear. Moreover, with the recent development of 3D sensors and 3D reconstruction techniques, a new research direction emerges to align cross-source point clouds. This survey conducts a comprehensive survey, including both same-source and cross-source registration methods, and summarize the connections between optimization-based and deep learning methods, to provide further research insight. This survey also builds a new benchmark to evaluate the state-of-the-art registration algorithms in solving cross-source challenges. Besides, this survey summarizes the benchmark data sets and discusses point cloud registration applications across various domains. Finally, this survey proposes potential research directions in this rapidly growing field.

preprint2021arXiv

An approximate dynamic programming approach to the admission control of elective patients

In this paper, we propose an approximate dynamic programming (ADP) algorithm to solve a Markov decision process (MDP) formulation for the admission control of elective patients. To manage the elective patients from multiple specialties equitably and efficiently, we establish a waiting list and assign each patient a time-dependent dynamic priority score. Then, taking the random arrivals of patients into account, sequential decisions are made on a weekly basis. At the end of each week, we select the patients to be treated in the following week from the waiting list. By minimizing the cost function of the MDP over an infinite horizon, we seek to achieve the best trade-off between the patients' waiting times and the over-utilization of surgical resources. Considering the curses of dimensionality resulting from the large scale of realistically sized problems, we first analyze the structural properties of the MDP and propose an algorithm that facilitates the search for best actions. We then develop a novel reinforcement-learning-based ADP algorithm as the solution technique. Experimental results reveal that the proposed algorithms consume much less computation time in comparison with that required by conventional dynamic programming methods. Additionally, the algorithms are shown to be capable of computing high-quality near-optimal policies for realistically sized problems.

preprint2021arXiv

Aurora Guard: Reliable Face Anti-Spoofing via Mobile Lighting System

Face authentication on mobile end has been widely applied in various scenarios. Despite the increasing reliability of cutting-edge face authentication/verification systems to variations like blinking eye and subtle facial expression, anti-spoofing against high-resolution rendering replay of paper photos or digital videos retains as an open problem. In this paper, we propose a simple yet effective face anti-spoofing system, termed Aurora Guard (AG). Our system firstly extracts the normal cues via light reflection analysis, and then adopts an end-to-end trainable multi-task Convolutional Neural Network (CNN) to accurately recover subjects' intrinsic depth and material map to assist liveness classification, along with the light CAPTCHA checking mechanism in the regression branch to further improve the system reliability. Experiments on public Replay-Attack and CASIA datasets demonstrate the merits of our proposed method over the state-of-the-arts. We also conduct extensive experiments on a large-scale dataset containing 12,000 live and diverse spoofing samples, which further validates the generalization ability of our method in the wild.

preprint2021arXiv

CSformer: Bridging Convolution and Transformer for Compressive Sensing

Convolution neural networks (CNNs) have succeeded in compressive image sensing. However, due to the inductive bias of locality and weight sharing, the convolution operations demonstrate the intrinsic limitations in modeling the long-range dependency. Transformer, designed initially as a sequence-to-sequence model, excels at capturing global contexts due to the self-attention-based architectures even though it may be equipped with limited localization abilities. This paper proposes CSformer, a hybrid framework that integrates the advantages of leveraging both detailed spatial information from CNN and the global context provided by transformer for enhanced representation learning. The proposed approach is an end-to-end compressive image sensing method, composed of adaptive sampling and recovery. In the sampling module, images are measured block-by-block by the learned sampling matrix. In the reconstruction stage, the measurement is projected into dual stems. One is the CNN stem for modeling the neighborhood relationships by convolution, and the other is the transformer stem for adopting global self-attention mechanism. The dual branches structure is concurrent, and the local features and global representations are fused under different resolutions to maximize the complementary of features. Furthermore, we explore a progressive strategy and window-based transformer block to reduce the parameter and computational complexity. The experimental results demonstrate the effectiveness of the dedicated transformer-based architecture for compressive sensing, which achieves superior performance compared to state-of-the-art methods on different datasets.

preprint2021arXiv

Exploiting Web Images for Fine-Grained Visual Recognition by Eliminating Noisy Samples and Utilizing Hard Ones

Labeling objects at a subordinate level typically requires expert knowledge, which is not always available when using random annotators. As such, learning directly from web images for fine-grained recognition has attracted broad attention. However, the presence of label noise and hard examples in web images are two obstacles for training robust fine-grained recognition models. Therefore, in this paper, we propose a novel approach for removing irrelevant samples from real-world web images during training, while employing useful hard examples to update the network. Thus, our approach can alleviate the harmful effects of irrelevant noisy web images and hard examples to achieve better performance. Extensive experiments on three commonly used fine-grained datasets demonstrate that our approach is far superior to current state-of-the-art web-supervised methods.

preprint2021arXiv

Revisiting BFloat16 Training

State-of-the-art generic low-precision training algorithms use a mix of 16-bit and 32-bit precision, creating the folklore that 16-bit hardware compute units alone are not enough to maximize model accuracy. As a result, deep learning accelerators are forced to support both 16-bit and 32-bit floating-point units (FPUs), which is more costly than only using 16-bit FPUs for hardware design. We ask: can we train deep learning models only with 16-bit floating-point units, while still matching the model accuracy attained by 32-bit training? Towards this end, we study 16-bit-FPU training on the widely adopted BFloat16 unit. While these units conventionally use nearest rounding to cast output to 16-bit precision, we show that nearest rounding for model weight updates often cancels small updates, which degrades the convergence and model accuracy. Motivated by this, we study two simple techniques well-established in numerical analysis, stochastic rounding and Kahan summation, to remedy the model accuracy degradation in 16-bit-FPU training. We demonstrate that these two techniques can enable up to 7% absolute validation accuracy gain in 16-bit-FPU training. This leads to 0.1% lower to 0.2% higher validation accuracy compared to 32-bit training across seven deep learning applications.

preprint2021arXiv

Semantically Meaningful Class Prototype Learning for One-Shot Image Semantic Segmentation

One-shot semantic image segmentation aims to segment the object regions for the novel class with only one annotated image. Recent works adopt the episodic training strategy to mimic the expected situation at testing time. However, these existing approaches simulate the test conditions too strictly during the training process, and thus cannot make full use of the given label information. Besides, these approaches mainly focus on the foreground-background target class segmentation setting. They only utilize binary mask labels for training. In this paper, we propose to leverage the multi-class label information during the episodic training. It will encourage the network to generate more semantically meaningful features for each category. After integrating the target class cues into the query features, we then propose a pyramid feature fusion module to mine the fused features for the final classifier. Furthermore, to take more advantage of the support image-mask pair, we propose a self-prototype guidance branch to support image segmentation. It can constrain the network for generating more compact features and a robust prototype for each semantic class. For inference, we propose a fused prototype guidance branch for the segmentation of the query image. Specifically, we leverage the prediction of the query image to extract the pseudo-prototype and combine it with the initial prototype. Then we utilize the fused prototype to guide the final segmentation of the query image. Extensive experiments demonstrate the superiority of our proposed approach.

preprint2021arXiv

Temporal-Amount Snapshot MultiGraph for Ethereum Transaction Tracking

With the wide application of blockchain in the financial field, the rise of various types of cybercrimes has brought great challenges to the security of blockchain. In order to better understand this emerging market and explore more efficient countermeasures for effective supervision, it is imperative to track transactions on blockchain-based systems. Due to the openness of Ethereum, we can easily access the publicly available transaction records, model them as a complex network, and further study the problem of transaction tracking via link prediction, which provides a deeper understanding of Ethereum transactions from a network perspective. Specifically, we introduce an embedding based link prediction framework that is composed of temporal-amount snapshot multigraph (TASMG) and present temporal-amount walk (TAW). By taking the realistic rules and features of transaction networks into consideration, we propose TASMG to model Ethereum transaction records as a temporal-amount network and then present TAW to effectively embed accounts via their transaction records, which integrates temporal and amount information of the proposed network. Experimental results demonstrate the superiority of the proposed framework in learning more informative representations and could be an effective method for transaction tracking.

preprint2020arXiv

A Similarity Inference Metric for RGB-Infrared Cross-Modality Person Re-identification

RGB-Infrared (IR) cross-modality person re-identification (re-ID), which aims to search an IR image in RGB gallery or vice versa, is a challenging task due to the large discrepancy between IR and RGB modalities. Existing methods address this challenge typically by aligning feature distributions or image styles across modalities, whereas the very useful similarities among gallery samples of the same modality (i.e. intra-modality sample similarities) is largely neglected. This paper presents a novel similarity inference metric (SIM) that exploits the intra-modality sample similarities to circumvent the cross-modality discrepancy targeting optimal cross-modality image matching. SIM works by successive similarity graph reasoning and mutual nearest-neighbor reasoning that mine cross-modality sample similarities by leveraging intra-modality sample similarities from two different perspectives. Extensive experiments over two cross-modality re-ID datasets (SYSU-MM01 and RegDB) show that SIM achieves significant accuracy improvement but with little extra training as compared with the state-of-the-art.

preprint2020arXiv

Automated Radiological Report Generation For Chest X-Rays With Weakly-Supervised End-to-End Deep Learning

The chest X-Ray (CXR) is the one of the most common clinical exam used to diagnose thoracic diseases and abnormalities. The volume of CXR scans generated daily in hospitals is huge. Therefore, an automated diagnosis system able to save the effort of doctors is of great value. At present, the applications of artificial intelligence in CXR diagnosis usually use pattern recognition to classify the scans. However, such methods rely on labeled databases, which are costly and usually have large error rates. In this work, we built a database containing more than 12,000 CXR scans and radiological reports, and developed a model based on deep convolutional neural network and recurrent network with attention mechanism. The model learns features from the CXR scans and the associated raw radiological reports directly; no additional labeling of the scans are needed. The model provides automated recognition of given scans and generation of reports. The quality of the generated reports was evaluated with both the CIDEr scores and by radiologists as well. The CIDEr scores are found to be around 5.8 on average for the testing dataset. Further blind evaluation suggested a comparable performance against human radiologist.

preprint2020arXiv

Computational prediction of RNA tertiary structures using machine learning methods

RNAs play crucial and versatile roles in biological processes. Computational prediction approaches can help to understand RNA structures and their stabilizing factors, thus providing information on their functions, and facilitating the design of new RNAs. Machine learning (ML) techniques have made tremendous progress in many fields in the past few years. Although their usage in protein-related fields has a long history, the use of ML methods in predicting RNA tertiary structures is new and rare. Here, we review the recent advances of using ML methods on RNA structure predictions and discuss the advantages and limitation, the difficulties and potentials of these approaches when applied in the field.

preprint2020arXiv

Contextual Embeddings: When Are They Worth It?

We study the settings for which deep contextual embeddings (e.g., BERT) give large improvements in performance relative to classic pretrained embeddings (e.g., GloVe), and an even simpler baseline---random word embeddings---focusing on the impact of the training set size and the linguistic properties of the task. Surprisingly, we find that both of these simpler baselines can match contextual embeddings on industry-scale data, and often perform within 5 to 10% accuracy (absolute) on benchmark tasks. Furthermore, we identify properties of data for which contextual embeddings give particularly large gains: language containing complex structure, ambiguous word usage, and words unseen in training.

preprint2020arXiv

Dynamical quantum determinants and Pfaffians

We introduce the dynamical quantum Pfaffian on the dynamical quantum general linear group and prove its fundamental transformation identity. Hyper quantum dynamical Pfaffian is also introduced and formulas connecting them are given.

preprint2020arXiv

DyNet: Dynamic Convolution for Accelerating Convolutional Neural Networks

Convolution operator is the core of convolutional neural networks (CNNs) and occupies the most computation cost. To make CNNs more efficient, many methods have been proposed to either design lightweight networks or compress models. Although some efficient network structures have been proposed, such as MobileNet or ShuffleNet, we find that there still exists redundant information between convolution kernels. To address this issue, we propose a novel dynamic convolution method to adaptively generate convolution kernels based on image contents. To demonstrate the effectiveness, we apply dynamic convolution on multiple state-of-the-art CNNs. On one hand, we can reduce the computation cost remarkably while maintaining the performance. For ShuffleNetV2/MobileNetV2/ResNet18/ResNet50, DyNet can reduce 37.0/54.7/67.2/71.3% FLOPs without loss of accuracy. On the other hand, the performance can be largely boosted if the computation cost is maintained. Based on the architecture MobileNetV3-Small/Large, DyNet achieves 70.3/77.1% Top-1 accuracy on ImageNet with an improvement of 2.9/1.9%. To verify the scalability, we also apply DyNet on segmentation task, the results show that DyNet can reduce 69.3% FLOPs while maintaining Mean IoU on segmentation task.

preprint2020arXiv

Face Anti-Spoofing Via Disentangled Representation Learning

Face anti-spoofing is crucial to security of face recognition systems. Previous approaches focus on developing discriminative models based on the features extracted from images, which may be still entangled between spoof patterns and real persons. In this paper, motivated by the disentangled representation learning, we propose a novel perspective of face anti-spoofing that disentangles the liveness features and content features from images, and the liveness features is further used for classification. We also put forward a Convolutional Neural Network (CNN) architecture with the process of disentanglement and combination of low-level and high-level supervision to improve the generalization capabilities. We evaluate our method on public benchmark datasets and extensive experimental results demonstrate the effectiveness of our method against the state-of-the-art competitors. Finally, we further visualize some results to help understand the effect and advantage of disentanglement.

preprint2020arXiv

Feature-metric Registration: A Fast Semi-supervised Approach for Robust Point Cloud Registration without Correspondences

We present a fast feature-metric point cloud registration framework, which enforces the optimisation of registration by minimising a feature-metric projection error without correspondences. The advantage of the feature-metric projection error is robust to noise, outliers and density difference in contrast to the geometric projection error. Besides, minimising the feature-metric projection error does not need to search the correspondences so that the optimisation speed is fast. The principle behind the proposed method is that the feature difference is smallest if point clouds are aligned very well. We train the proposed method in a semi-supervised or unsupervised approach, which requires limited or no registration label data. Experiments demonstrate our method obtains higher accuracy and robustness than the state-of-the-art methods. Besides, experimental results show that the proposed method can handle significant noise and density difference, and solve both same-source and cross-source point cloud registration.

preprint2020arXiv

Gelfand-Tsetlin modules for gl(n,m)

We address the problem of classifying of irreducible Gelfand-Tsetlin modules for gl(m|n) and show that it reduces to the classification of Gelfand-Tsetlin modules for the even part. We also give an explicit tableaux construction and the irreducibility criterion for the class of quasi typical and quasi covariant Gelfand-Tsetlin modules which includes all essentially typical and covariant tensor finite dimensional modules. In the quasi typical case new irreducible representations are infinite dimensional gl(m|n)-modules which are isomorphic to the parabolically induced (Kac) modules.

preprint2020arXiv

GID-Net: Detecting Human-Object Interaction with Global and Instance Dependency

Since detecting and recognizing individual human or object are not adequate to understand the visual world, learning how humans interact with surrounding objects becomes a core technology. However, convolution operations are weak in depicting visual interactions between the instances since they only build blocks that process one local neighborhood at a time. To address this problem, we learn from human perception in observing HOIs to introduce a two-stage trainable reasoning mechanism, referred to as GID block. GID block breaks through the local neighborhoods and captures long-range dependency of pixels both in global-level and instance-level from the scene to help detecting interactions between instances. Furthermore, we conduct a multi-stream network called GID-Net, which is a human-object interaction detection framework consisting of a human branch, an object branch and an interaction branch. Semantic information in global-level and local-level are efficiently reasoned and aggregated in each of the branches. We have compared our proposed GID-Net with existing state-of-the-art methods on two public benchmarks, including V-COCO and HICO-DET. The results have showed that GID-Net outperforms the existing best-performing methods on both the above two benchmarks, validating its efficacy in detecting human-object interactions.

preprint2020arXiv

Glioma Grade Prediction Using Wavelet Scattering-Based Radiomics

Glioma grading before surgery is very critical for the prognosis prediction and treatment plan making. We present a novel wavelet scattering-based radiomic method to predict noninvasively and accurately the glioma grades. The method consists of wavelet scattering feature extraction, dimensionality reduction, and glioma grade prediction. The dimensionality reduction was achieved using partial least squares (PLS) regression and the glioma grade prediction using support vector machine (SVM), logistic regression (LR) and random forest (RF). The prediction obtained on multimodal magnetic resonance images of 285 patients with well-labeled intratumoral and peritumoral regions showed that the area under the receiver operating characteristic curve (AUC) of glioma grade prediction was increased up to 0.99 when considering both intratumoral and peritumoral features in multimodal images, which represents an increase of about 13% compared to traditional radiomics. In addition, the features extracted from peritumoral regions further increase the accuracy of glioma grading.

preprint2020arXiv

Hardware Accelerator for Adversarial Attacks on Deep Learning Neural Networks

Recent studies identify that Deep learning Neural Networks (DNNs) are vulnerable to subtle perturbations, which are not perceptible to human visual system but can fool the DNN models and lead to wrong outputs. A class of adversarial attack network algorithms has been proposed to generate robust physical perturbations under different circumstances. These algorithms are the first efforts to move forward secure deep learning by providing an avenue to train future defense networks, however, the intrinsic complexity of them prevents their broader usage. In this paper, we propose the first hardware accelerator for adversarial attacks based on memristor crossbar arrays. Our design significantly improves the throughput of a visual adversarial perturbation system, which can further improve the robustness and security of future deep learning systems. Based on the algorithm uniqueness, we propose four implementations for the adversarial attack accelerator ($A^3$) to improve the throughput, energy efficiency, and computational efficiency.

preprint2020arXiv

Imprint of a new light particle at KOTO?

Recently, the KOTO experiment reported their new preliminary result of searching for the decay $K_L\toπ^0ν\barν$. Three candidate events were observed in the signal region, which exceed significantly the expectation based on the standard model. On the other hand, the new NA62 and previous BNL-E949 experiments yielded a consistent result and confirmed the standard model prediction in the charged meson decay $K^+\toπ^+ν\barν$. Furthermore, the two decays are bound by a well-motivated relation from an analysis of isospin symmetry that is hard to break by the new physics of heavy particles. In this work, we study the issue by a systematic effective field theory approach with three of the simplest scenarios, in which the $K_L$ may decay into a new light neutral particle $X$, i.e., $K_L\toπ^0X$, $K_L\to γγX$, or $K_L\toπ^0XX$. We assess the feasibility of the scenarios by simulations and by incorporating constraints coming from NA62 and other relevant experiments. Our main conclusion is that the scenario $K\toπXX$ for a long lived scalar $X$ seems more credible than the other two when combining distributions and other experimental constraints while the region below the KOTO's blind box provides a good detection environment to search for all three scenarios for a relatively heavy $X$.

preprint2020arXiv

Max-min Fairness of K-user Cooperative Rate-Splitting in MISO Broadcast Channel with User Relaying

Cooperative Rate-Splitting (CRS) strategy, relying on linearly precoded rate-splitting at the transmitter and opportunistic transmission of the common message by the relaying user, has recently been shown to outperform typical Non-cooperative Rate-Splitting (NRS), Cooperative Non-Orthogonal Multiple Access (C-NOMA) and Space Division Multiple Access (SDMA) in a two-user Multiple Input Single Output (MISO) Broadcast Channel (BC) with user relaying. In this work, the existing two-user CRS transmission strategy is generalized to the K-user case. We study the problem of jointly optimizing the precoders, message split, time slot allocation, and relaying user scheduling with the objective of maximizing the minimum rate among users. An efficient self-organizing relaying protocol is first proposed followed by a Successive Convex Approximation (SCA)-based algorithm to jointly optimize time slot, precoders and message split. Numerical results show that the worst-case achievable rate achieved by CRS is significantly increased over that of NRS and SDMA in a wide range of network loads and user deployments. Importantly, the proposed SCA-based algorithm dramatically reduces the computational complexity without any rate loss compared with the conventional algorithm in the literature of CRS. Therefore, we conclude that the proposed K-user CRS is more powerful than the existing transmission schemes.

preprint2020arXiv

Multi-factorial Optimization for Large-scale Virtual Machine Placement in Cloud Computing

The placement scheme of virtual machines (VMs) to physical servers (PSs) is crucial to lowering operational cost for cloud providers. Evolutionary algorithms (EAs) have been performed promising-solving on virtual machine placement (VMP) problems in the past. However, as growing demand for cloud services, the existing EAs fail to implement in large-scale virtual machine placement (LVMP) problem due to the high time complexity and poor scalability. Recently, the multi-factorial optimization (MFO) technology has surfaced as a new search paradigm in evolutionary computing. It offers the ability to evolve multiple optimization tasks simultaneously during the evolutionary process. This paper aims to apply the MFO technology to the LVMP problem in heterogeneous environment. Firstly, we formulate a deployment cost based VMP problem in the form of the MFO problem. Then, a multi-factorial evolutionary algorithm (MFEA) embedded with greedy-based allocation operator is developed to address the established MFO problem. After that, a re-migration and merge operator is designed to offer the integrated solution of the LVMP problem from the solutions of MFO problem. To assess the effectiveness of our proposed method, the simulation experiments are carried on large-scale and extra large-scale VMs test data sets. The results show that compared with various heuristic methods, our method could shorten optimization time significantly and offer a competitive placement solution for the LVMP problem in heterogeneous environment.

preprint2020arXiv

On the Downstream Performance of Compressed Word Embeddings

Compressing word embeddings is important for deploying NLP models in memory-constrained settings. However, understanding what makes compressed embeddings perform well on downstream tasks is challenging---existing measures of compression quality often fail to distinguish between embeddings that perform well and those that do not. We thus propose the eigenspace overlap score as a new measure. We relate the eigenspace overlap score to downstream performance by developing generalization bounds for the compressed embeddings in terms of this score, in the context of linear and logistic regression. We then show that we can lower bound the eigenspace overlap score for a simple uniform quantization compression method, helping to explain the strong empirical performance of this method. Finally, we show that by using the eigenspace overlap score as a selection criterion between embeddings drawn from a representative set we compressed, we can efficiently identify the better performing embedding with up to $2\times$ lower selection error rates than the next best measure of compression quality, and avoid the cost of training a model for each task of interest.

preprint2020arXiv

Persistent spin dynamics and absence of spin freezing in the $H$-$T$ phase diagram of the 2D triangular antiferromagnet YbMgGaO$_4$

We report results of muon spin relaxation and rotation ($μ$SR) experiments on the spin-liquid candidate~YbMgGaO$_{4}$. No static magnetism $\gtrsim 0.003μ_B$ per Yb ion, ordered or disordered, is observed down to 22~mK, a factor of two lower in temperature than previous measurements. Persistent (temperature-independent) spin dynamics are observed up to 0.20~K and at least 1~kOe, thus extending previous zero-field $μ$SR results over a substantial region of the $H$-$T$ phase diagram. Knight shift measurements in a 10-kOe transverse field reveal two lines with nearly equal amplitudes. Inhomogeneous muon depolarization in a longitudinal field, previously characterized by stretched-exponential relaxation due to spatial inhomogeneity, is fit equally well with two exponentials, also of equal amplitudes. We attribute these results to two interstitial muon sites in the unit cell, rather than disorder or other spatial distribution. Further evidence for this attribution is found from agreement between the ratio of the two measured relaxation rates and calculated mean-square local Yb$^{3+}$ dipolar fields at candidate muon sites. Zero-field data can be understood as a combination of two-exponential dynamic relaxation and quasistatic nuclear dipolar fields.

preprint2020arXiv

PipeMare: Asynchronous Pipeline Parallel DNN Training

Pipeline parallelism (PP) when training neural networks enables larger models to be partitioned spatially, leading to both lower network communication and overall higher hardware utilization. Unfortunately, to preserve the statistical efficiency of sequential training, existing PP techniques sacrifice hardware efficiency by decreasing pipeline utilization or incurring extra memory costs. In this paper, we investigate to what extent these sacrifices are necessary. We devise PipeMare, a simple yet robust training method that tolerates asynchronous updates during PP execution without sacrificing utilization or memory, which allows efficient use of fine-grained pipeline parallelism. Concretely, when tested on ResNet and Transformer networks, asynchrony enables PipeMare to use up to $2.7\times$ less memory or get $4.3\times$ higher pipeline utilization, with similar model quality, when compared to state-of-the-art synchronous PP training techniques.

preprint2020arXiv

Salvage Reusable Samples from Noisy Data for Robust Learning

Due to the existence of label noise in web images and the high memorization capacity of deep neural networks, training deep fine-grained (FG) models directly through web images tends to have an inferior recognition ability. In the literature, to alleviate this issue, loss correction methods try to estimate the noise transition matrix, but the inevitable false correction would cause severe accumulated errors. Sample selection methods identify clean ("easy") samples based on the fact that small losses can alleviate the accumulated errors. However, "hard" and mislabeled examples that can both boost the robustness of FG models are also dropped. To this end, we propose a certainty-based reusable sample selection and correction approach, termed as CRSSC, for coping with label noise in training deep FG models with web images. Our key idea is to additionally identify and correct reusable samples, and then leverage them together with clean examples to update the networks. We demonstrate the superiority of the proposed approach from both theoretical and experimental perspectives.

preprint2020arXiv

Scalar Coupling Constant Prediction Using Graph Embedding Local Attention Encoder

Scalar coupling constant (SCC) plays a key role in the analysis of three-dimensional structure of organic matter, however, the traditional SCC prediction using quantum mechanical calculations is very time-consuming. To calculate SCC efficiently and accurately, we proposed a graph embedding local self-attention encoder (GELAE) model, in which, a novel invariant structure representation of the coupling system in terms of bond length, bond angle and dihedral angle was presented firstly, and then a local self-attention module embedded with the adjacent matrix of a graph was designed to extract effectively the features of coupling systems, finally, with a modified classification loss function, the SCC was predicted. To validate the superiority of the proposed method, we conducted a series of comparison experiments using different structure representations, different attention modules, and different losses. The experimental results demonstrate that, compared to the traditional chemical bond structure representations, the rotation and translation invariant structure representations proposed in this work can improve the SCC prediction accuracy; with the graph embedded local self-attention, the mean absolute error (MAE) of the prediction model in the validation set decreases from 0.1603 Hz to 0.1067 Hz; using the classification based loss function instead of the scaled regression loss, the MAE of the predicted SCC can be decreased to 0.0963 HZ, which is close to the quantum chemistry standard on CHAMPS dataset.

preprint2020arXiv

Stochastic Batch Augmentation with An Effective Distilled Dynamic Soft Label Regularizer

Data augmentation have been intensively used in training deep neural network to improve the generalization, whether in original space (e.g., image space) or representation space. Although being successful, the connection between the synthesized data and the original data is largely ignored in training, without considering the distribution information that the synthesized samples are surrounding the original sample in training. Hence, the behavior of the network is not optimized for this. However, that behavior is crucially important for generalization, even in the adversarial setting, for the safety of the deep learning system. In this work, we propose a framework called Stochastic Batch Augmentation (SBA) to address these problems. SBA stochastically decides whether to augment at iterations controlled by the batch scheduler and in which a ''distilled'' dynamic soft label regularization is introduced by incorporating the similarity in the vicinity distribution respect to raw samples. The proposed regularization provides direct supervision by the KL-Divergence between the output soft-max distributions of original and virtual data. Our experiments on CIFAR-10, CIFAR-100, and ImageNet show that SBA can improve the generalization of the neural networks and speed up the convergence of network training.

preprint2020arXiv

Time-aware Gradient Attack on Dynamic Network Link Prediction

In network link prediction, it is possible to hide a target link from being predicted with a small perturbation on network structure. This observation may be exploited in many real world scenarios, for example, to preserve privacy, or to exploit financial security. There have been many recent studies to generate adversarial examples to mislead deep learning models on graph data. However, none of the previous work has considered the dynamic nature of real-world systems. In this work, we present the first study of adversarial attack on dynamic network link prediction (DNLP). The proposed attack method, namely time-aware gradient attack (TGA), utilizes the gradient information generated by deep dynamic network embedding (DDNE) across different snapshots to rewire a few links, so as to make DDNE fail to predict target links. We implement TGA in two ways: one is based on traversal search, namely TGA-Tra; and the other is simplified with greedy search for efficiency, namely TGA-Gre. We conduct comprehensive experiments which show the outstanding performance of TGA in attacking DNLP algorithms.

preprint2020arXiv

Understanding Graph Neural Networks from Graph Signal Denoising Perspectives

Graph neural networks (GNNs) have attracted much attention because of their excellent performance on tasks such as node classification. However, there is inadequate understanding on how and why GNNs work, especially for node representation learning. This paper aims to provide a theoretical framework to understand GNNs, specifically, spectral graph convolutional networks and graph attention networks, from graph signal denoising perspectives. Our framework shows that GNNs are implicitly solving graph signal denoising problems: spectral graph convolutions work as denoising node features, while graph attentions work as denoising edge weights. We also show that a linear self-attention mechanism is able to compete with the state-of-the-art graph attention methods. Our theoretical results further lead to two new models, GSDN-F and GSDN-EF, which work effectively for graphs with noisy node features and/or noisy edges. We validate our theoretical findings and also the effectiveness of our new models by experiments on benchmark datasets. The source code is available at \url{https://github.com/fuguoji/GSDN}.

preprint2020arXiv

Understanding the Downstream Instability of Word Embeddings

Many industrial machine learning (ML) systems require frequent retraining to keep up-to-date with constantly changing data. This retraining exacerbates a large challenge facing ML systems today: model training is unstable, i.e., small changes in training data can cause significant changes in the model's predictions. In this paper, we work on developing a deeper understanding of this instability, with a focus on how a core building block of modern natural language processing (NLP) pipelines---pre-trained word embeddings---affects the instability of downstream NLP models. We first empirically reveal a tradeoff between stability and memory: increasing the embedding memory 2x can reduce the disagreement in predictions due to small changes in training data by 5% to 37% (relative). To theoretically explain this tradeoff, we introduce a new measure of embedding instability---the eigenspace instability measure---which we prove bounds the disagreement in downstream predictions introduced by the change in word embeddings. Practically, we show that the eigenspace instability measure can be a cost-effective way to choose embedding parameters to minimize instability without training downstream models, outperforming other embedding distance measures and performing competitively with a nearest neighbor-based measure. Finally, we demonstrate that the observed stability-memory tradeoffs extend to other types of embeddings as well, including knowledge graph and contextual word embeddings.

preprint2020arXiv

Weyl Superconductivity in UTe2

The search for a material platform for topological quantum computation has recently focused on unconventional superconductors. Such material systems, where the superconducting order parameter breaks a symmetry of the crystal point group, are capable of hosting novel phenomena, including emergent Majorana quasiparticles. Unique among unconventional superconductors is the recently discovered UTe2, where spin-triplet superconductivity emerges from a paramagnetic normal state. Although UTe2 could be considered a relative of a family of known ferromagnetic superconductors, the unique crystal structure of this material and experimentally suggested zero Curie temperature pose a great challenge to determining the symmetries, magnetism, and topology underlying the superconducting state. These emergent properties will determine the utility of UTe2 for future spintronics and quantum information applications. Here, we report observations of a non-zero polar Kerr effect and of two transitions in the specific heat upon entering the superconducting state, which together show that the superconductivity in UTe2 is characterized by an order parameter with two components that breaks time reversal symmetry. These data allow us to place firm constraints on the symmetries of the order parameter, which strongly suggest that UTe2 is a Weyl superconductor that hosts chiral Fermi arc surface states.

preprint2019arXiv

Compare More Nuanced:Pairwise Alignment Bilinear Network For Few-shot Fine-grained Learning

The recognition ability of human beings is developed in a progressive way. Usually, children learn to discriminate various objects from coarse to fine-grained with limited supervision. Inspired by this learning process, we propose a simple yet effective model for the Few-Shot Fine-Grained (FSFG) recognition, which tries to tackle the challenging fine-grained recognition task using meta-learning. The proposed method, named Pairwise Alignment Bilinear Network (PABN), is an end-to-end deep neural network. Unlike traditional deep bilinear networks for fine-grained classification, which adopt the self-bilinear pooling to capture the subtle features of images, the proposed model uses a novel pairwise bilinear pooling to compare the nuanced differences between base images and query images for learning a deep distance metric. In order to match base image features with query image features, we design feature alignment losses before the proposed pairwise bilinear pooling. Experiment results on four fine-grained classification datasets and one generic few-shot dataset demonstrate that the proposed model outperforms both the state-ofthe-art few-shot fine-grained and general few-shot methods.

preprint2019arXiv

Cooperative Rate-Splitting for MISO Broadcast Channel with User Relaying, and Performance Benefits over Cooperative NOMA

Due to its promising performance in a wide range of practical scenarios, Rate-Splitting (RS) has recently received significant attention in academia for the downlink of communication systems. In this letter, we propose and analyse a Cooperative Rate-Splitting (CRS) strategy based on the three-node relay channel where the transmitter is equipped with multiple antennas. By splitting user messages and linearly precoding common and private streams at the transmitter, and opportunistically asking the relaying user to forward its decoded common message, CRS can efficiently cope with a wide range of propagation conditions (disparity of user channel strengths and directions) and compensate for the performance degradation due to deep fading. The precoder design and the resource allocation are optimized by solving the Weighted Sum Rate (WSR) maximization problem. Numerical results demonstrate that our proposed CRS scheme can achieve an explicit rate region improvement compared to its non-cooperative counterpart and other cooperative strategies (such as cooperative NOMA).

preprint2019arXiv

Deep levels analysis in wavelength extended InGaAsBi photodetector

InP based dilute Bismide InGaAsBi material is emerging as a promising candidate for extending short wavelength infrared detection. One critical factor to limit the performance of these InGaAsBi photodiodes is dark current caused by defects within the material. In this work, low frequency noise spectroscopy (LFNS) and temperature varied photoluminescence was used to characterize the defect levels in the devices. Three deep levels located at Ec -0.33 eV, Ev +0.14 eV, and Ec -0.51 eV were identified from the LFNS spectra, which are consistent with emission peak energy found by photoluminescence spectra of InGaAsBi.

preprint2019arXiv

Feasibility study of TPC tracker detector for the circular collider

The discovery of a SM Higgs boson at the LHC brought about great opportunity to investigate the feasibility of a Circular Electron Positron Collider (CEPC) operating at center-of-mass energy of $\sim 240$ GeV, as a Higgs factory, with designed luminosity of about $2\times 10^{34}cm^{-2}s^{-1}$. The CEPC provides a much cleaner collision environment than the LHC, it is ideally suited for studying the properties of Higgs boson with greater precision. Another advantage of the CEPC over the LHC is that the Higgs boson can be detected through the recoil mass method by only reconstructing Z boson decay without examining the Higgs decays. In Concept Design Report(CDR), the circumference of CEPC is 100km, with two interaction points available for exploring different detector design scenarios and technologies. The baseline design of CEPC detector is an ILD-like concept, with a superconducting solenoid of 3.0 Tesla surrounding the inner silicon detector, TPC tracker detector and the calorimetry system. Time Projection Chambers (TPCs) have been extensively studied and used in many fields, especially in particle physics experiments, including STAR and ALICE. The TPC detector will operate in continuous mode on the circular machine. To fulfill the physics goals of the future circular collider and meet Higgs/$Z$ run, a TPC with excellent performance is required. We have proposed and investigated the ions controlling performance of a novel configuration detector module. The aim of this study is to suppress ion backflow ($IBF$) continually. In this paper, some update results of the feasibility and limitation on TPC detector technology R$\&$D will be given using the hybrid gaseous detector module.

preprint2019arXiv

PMC-GANs: Generating Multi-Scale High-Quality Pedestrian with Multimodal Cascaded GANs

Recently, generative adversarial networks (GANs) have shown great advantages in synthesizing images, leading to a boost of explorations of using faked images to augment data. This paper proposes a multimodal cascaded generative adversarial networks (PMC-GANs) to generate realistic and diversified pedestrian images and augment pedestrian detection data. The generator of our model applies a residual U-net structure, with multi-scale residual blocks to encode features, and attention residual blocks to help decode and rebuild pedestrian images. The model constructs in a coarse-to-fine fashion and adopts cascade structure, which is beneficial to produce high-resolution pedestrians. PMC-GANs outperforms baselines, and when used for data augmentation, it improves pedestrian detection results.

preprint2018arXiv

Structured Control Nets for Deep Reinforcement Learning

In recent years, Deep Reinforcement Learning has made impressive advances in solving several important benchmark problems for sequential decision making. Many control applications use a generic multilayer perceptron (MLP) for non-vision parts of the policy network. In this work, we propose a new neural network architecture for the policy network representation that is simple yet effective. The proposed Structured Control Net (SCN) splits the generic MLP into two separate sub-modules: a nonlinear control module and a linear control module. Intuitively, the nonlinear control is for forward-looking and global control, while the linear control stabilizes the local dynamics around the residual of global control. We hypothesize that this will bring together the benefits of both linear and nonlinear policies: improve training sample efficiency, final episodic reward, and generalization of learned policy, while requiring a smaller network and being generally applicable to different training methods. We validated our hypothesis with competitive results on simulations from OpenAI MuJoCo, Roboschool, Atari, and a custom 2D urban driving environment, with various ablation and generalization tests, trained with multiple black-box and policy gradient training methods. The proposed architecture has the potential to improve upon broader control tasks by incorporating problem specific priors into the architecture. As a case study, we demonstrate much improved performance for locomotion tasks by emulating the biological central pattern generators (CPGs) as the nonlinear part of the architecture.

preprint2014arXiv

Quantum Pfaffians and Hyper-Pfaffians

The concept of the quantum Pfaffian is rigorously examined and refurbished using the new method of quantum exterior algebras. We derive a complete family of Plücker relations for the quantum linear transformations, and then use them to give an optimal set of relations required for the quantum Pfaffian. We then give the formula between the quantum determinant and the quantum Pfaffian and prove that any quantum determinant can be expressed as a quantum Pfaffian. Finally the quantum hyper-Pfaffian is introduced, and we prove a similar result of expressing quantum determinants in terms of quantum hyper-Pfaffians at modular cases.

Institution

Affiliation not imported yet

This author record came from a source that does not expose affiliation metadata. Once the author claims the profile or we enrich the record from another provider, this section will link to the concrete institution.

Source provenance

Where this author record came from

arxivconfidence 95%

external id: arxiv:2605.05813:author:3:jian-zhang

Imported May 20, 2026Synced May 21, 2026

arxivconfidence 95%

external id: arxiv:2605.18224:author:3:jian-zhang

Imported May 20, 2026Synced May 21, 2026

arxivconfidence 95%

external id: arxiv:2605.19852:author:4:jian-zhang

Imported May 20, 2026Synced May 21, 2026

arxivconfidence 95%

external id: arxiv:2605.02134:author:6:jian-zhang