BZPEERReading, trust and collaboration for research
Menu

Discover

GraphFollow connections around a paper, topic or saved view.

Workspaces

HomePick up where your research flow left off.InboxHandle requests, replies and notifications from one place.CollectionsCurate reading lists, evidence sets and shareable dossiers.ResearchSee your private provenance graph, trail and imports.CollaborationMove from discovery into focused discussion rooms.PublishDraft, preview and publish full-text research articles.

Network

ExpertsFind specialists worth following or asking for help.CommunitiesJoin research circles organized around shared work.PartnersSee external organizations active around the same topics.

Opportunities

ListingsBrowse roles, calls and collaborations with clearer fit signals.

Account

Log inReturn to your reading and collaboration workspace.Create accountStart saving work, following people and joining teams.
Log inCreate account

Researcher profile

Jiacheng Li

Jiacheng Li contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
Computer VisionMachine LearningComputation and LanguageBiomoleculesArtificial IntelligenceInformation RetrievalCell BehaviorCryptography and Securityeess.IVSocial and Information Networks

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
14works
0followers
10topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

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

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

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.

Published work

14 published item(s)

preprint2026arXiv

AFA-LoRA: Enabling Non-Linear Adaptations in LoRA with Activation Function Annealing

Low-Rank Adaptation (LoRA) is a widely adopted parameter-efficient fine-tuning (PEFT) method. However, its linear adaptation process limits its expressive power. This means there is a gap between the expressive power of linear training and non-linear training. To bridge this gap, we propose AFA-LoRA, a novel training strategy that brings non-linear expressivity to LoRA while maintaining its seamless mergeability. Our key innovation is an annealed activation function that transitions from a non-linear to a linear transformation during training, allowing the adapter to initially adopt stronger representational capabilities before converging to a mergeable linear form. We implement our method on supervised fine-tuning, reinforcement learning, and speculative decoding. The results show that AFA-LoRA reduces the performance gap between LoRA and full-parameter training. This work enables a more powerful and practical paradigm of parameter-efficient adaptation.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Generating HDR Video from SDR Video

The high dynamic range (HDR) video ecosystem is approaching maturity, but the problem of upconverting legacy standard dynamic range (SDR) videos persists without a convincing solution. We propose a framework for HDR video synthesis from casual SDR footage by leveraging large-scale generative video models. We introduce a Multi-Exposure Video Model (MEVM) that can predict exposure-bracketed linear SDR video sequences from a single nonlinear SDR video input. We further propose a learnable Video Merging Model (VMM) that merges the predicted exposure-bracketed video into a high-quality HDR sequence while preserving detail in both shadows and highlights. Extensive experiments, quantitative and qualitative evaluation, and a user study demonstrate that our approach enables robust HDR conversion for in-the-wild examples from casual consumer videos and even iconic films. Finally, our model can support HDR synthesis pipelines built upon existing SDR generative video models. Output HDR videos can be viewed on our supplementary webpage: sdr2hdrvideo.github.io

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

MARBLE: Multi-Aspect Reward Balance for Diffusion RL

Reinforcement learning fine-tuning has become the dominant approach for aligning diffusion models with human preferences. However, assessing images is intrinsically a multi-dimensional task, and multiple evaluation criteria need to be optimized simultaneously. Existing practice deal with multiple rewards by training one specialist model per reward, optimizing a weighted-sum reward $R(x)=\sum_k w_k R_k(x)$, or sequentially fine-tuning with a hand-crafted stage schedule. These approaches either fail to produce a unified model that can be jointly trained on all rewards or necessitates heavy manually tuned sequential training. We find that the failure stems from using a naive weighted-sum reward aggregation. This approach suffers from a sample-level mismatch because most rollouts are specialist samples, highly informative for certain reward dimensions but irrelevant for others; consequently, weighted summation dilutes their supervision. To address this issue, we propose MARBLE (Multi-Aspect Reward BaLancE), a gradient-space optimization framework that maintains independent advantage estimators for each reward, computes per-reward policy gradients, and harmonizes them into a single update direction without manually-tuned reward weighting, by solving a Quadratic Programming problem. We further propose an amortized formulation that exploits the affine structure of the loss used in DiffusionNFT, to reduce the per-step cost from K+1 backward passes to near single-reward baseline cost, together with EMA smoothing on the balancing coefficients to stabilize updates against transient single-batch fluctuations. On SD3.5 Medium with five rewards, MARBLE improves all five reward dimensions simultaneously, turns the worst-aligned reward's gradient cosine from negative under weighted summation in 80% of mini-batches to consistently positive, and runs at 0.97X the training speed of baseline training.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

SpatialFusion: Endowing Unified Image Generation with Intrinsic 3D Geometric Awareness

Recent unified image generation models have achieved remarkable success by employing MLLMs for semantic understanding and diffusion backbones for image generation. However, these models remain fundamentally limited in spatially-aware tasks due to a lack of intrinsic spatial understanding and the absence of explicit geometric guidance during generation. In this paper, we propose SpatialFusion, a novel framework that internalizes 3D geometric awareness into unified image generation models. Specifically, we first employ a Mixture-of-Transformers (MoT) architecture to augment the MLLM with a parallel spatial transformer to enhance 3D geometric modeling capability. By sharing self-attention with the MLLM, the spatial transformer learns to derive metric-depth maps of target images from rich semantic contexts. These explicit geometric scaffolds are then injected into the diffusion backbone through a specialized depth adapter, providing precise spatial constraints for spatially-coherent image generation. Through a progressive two-stage training strategy, SpatialFusion significantly enhances performance on spatially-aware benchmarks, notably outperforming leading models such as GPT-4o. Additionally, it achieves generalized performance gains across both text-to-image generation and image editing scenarios, all while maintaining negligible inference overhead.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Coarse-to-Fine Sparse Sequential Recommendation

Sequential recommendation aims to model dynamic user behavior from historical interactions. Self-attentive methods have proven effective at capturing short-term dynamics and long-term preferences. Despite their success, these approaches still struggle to model sparse data, on which they struggle to learn high-quality item representations. We propose to model user dynamics from shopping intents and interacted items simultaneously. The learned intents are coarse-grained and work as prior knowledge for item recommendation. To this end, we present a coarse-to-fine self-attention framework, namely CaFe, which explicitly learns coarse-grained and fine-grained sequential dynamics. Specifically, CaFe first learns intents from coarse-grained sequences which are dense and hence provide high-quality user intent representations. Then, CaFe fuses intent representations into item encoder outputs to obtain improved item representations. Finally, we infer recommended items based on representations of items and corresponding intents. Experiments on sparse datasets show that CaFe outperforms state-of-the-art self-attentive recommenders by 44.03% NDCG@5 on average.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

Gray Matter Segmentation in Ultra High Resolution 7 Tesla ex vivo T2w MRI of Human Brain Hemispheres

Ex vivo MRI of the brain provides remarkable advantages over in vivo MRI for visualizing and characterizing detailed neuroanatomy. However, automated cortical segmentation methods in ex vivo MRI are not well developed, primarily due to limited availability of labeled datasets, and heterogeneity in scanner hardware and acquisition protocols. In this work, we present a high resolution 7 Tesla dataset of 32 ex vivo human brain specimens. We benchmark the cortical mantle segmentation performance of nine neural network architectures, trained and evaluated using manually-segmented 3D patches sampled from specific cortical regions, and show excellent generalizing capabilities across whole brain hemispheres in different specimens, and also on unseen images acquired at different magnetic field strength and imaging sequences. Finally, we provide cortical thickness measurements across key regions in 3D ex vivo human brain images. Our code and processed datasets are publicly available at https://github.com/Pulkit-Khandelwal/picsl-ex-vivo-segmentation.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

MuCGEC: a Multi-Reference Multi-Source Evaluation Dataset for Chinese Grammatical Error Correction

This paper presents MuCGEC, a multi-reference multi-source evaluation dataset for Chinese Grammatical Error Correction (CGEC), consisting of 7,063 sentences collected from three Chinese-as-a-Second-Language (CSL) learner sources. Each sentence is corrected by three annotators, and their corrections are carefully reviewed by a senior annotator, resulting in 2.3 references per sentence. We conduct experiments with two mainstream CGEC models, i.e., the sequence-to-sequence model and the sequence-to-edit model, both enhanced with large pretrained language models, achieving competitive benchmark performance on previous and our datasets. We also discuss CGEC evaluation methodologies, including the effect of multiple references and using a char-based metric. Our annotation guidelines, data, and code are available at \url{https://github.com/HillZhang1999/MuCGEC}.

0 citations0 reviewsNo code linkedOpen access
preprint2022arXiv

UCTopic: Unsupervised Contrastive Learning for Phrase Representations and Topic Mining

High-quality phrase representations are essential to finding topics and related terms in documents (a.k.a. topic mining). Existing phrase representation learning methods either simply combine unigram representations in a context-free manner or rely on extensive annotations to learn context-aware knowledge. In this paper, we propose UCTopic, a novel unsupervised contrastive learning framework for context-aware phrase representations and topic mining. UCTopic is pretrained in a large scale to distinguish if the contexts of two phrase mentions have the same semantics. The key to pretraining is positive pair construction from our phrase-oriented assumptions. However, we find traditional in-batch negatives cause performance decay when finetuning on a dataset with small topic numbers. Hence, we propose cluster-assisted contrastive learning(CCL) which largely reduces noisy negatives by selecting negatives from clusters and further improves phrase representations for topics accordingly. UCTopic outperforms the state-of-the-art phrase representation model by 38.2% NMI in average on four entity cluster-ing tasks. Comprehensive evaluation on topic mining shows that UCTopic can extract coherent and diverse topical phrases.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Hydrophobic interaction determines docking affinity of SARS CoV 2 variants with antibodies

Preliminary epidemiologic, phylogenetic and clinical findings suggest that several novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants have increased transmissibility and decreased efficacy of several existing vaccines. Four mutations in the receptor-binding domain (RBD) of the spike protein that are reported to contribute to increased transmission. Understanding physical mechanism responsible for the affinity enhancement between the SARS-CoV-2 variants and ACE2 is the "urgent challenge" for developing blockers, vaccines and therapeutic antibodies against the coronavirus disease 2019 (COVID-19) pandemic. Based on a hydrophobic-interaction-based protein docking mechanism, this study reveals that the mutation N501Y obviously increased the hydrophobic attraction and decrease hydrophilic repulsion between the RBD and ACE2 that most likely caused the transmissibility increment of the variants. By analyzing the mutation-induced hydrophobic surface changes in the attraction and repulsion at the binding site of the complexes of the SARS-CoV-2 variants and antibodies, we found out that all the mutations of N501Y, E484K, K417N and L452R can selectively decrease or increase their binding affinity with some antibodies.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Membership Inference Attacks and Defenses in Classification Models

We study the membership inference (MI) attack against classifiers, where the attacker's goal is to determine whether a data instance was used for training the classifier. Through systematic cataloging of existing MI attacks and extensive experimental evaluations of them, we find that a model's vulnerability to MI attacks is tightly related to the generalization gap -- the difference between training accuracy and test accuracy. We then propose a defense against MI attacks that aims to close the gap by intentionally reduces the training accuracy. More specifically, the training process attempts to match the training and validation accuracies, by means of a new {\em set regularizer} using the Maximum Mean Discrepancy between the softmax output empirical distributions of the training and validation sets. Our experimental results show that combining this approach with another simple defense (mix-up training) significantly improves state-of-the-art defense against MI attacks, with minimal impact on testing accuracy.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

A hydrophobic-interaction-based mechanism trigger docking between the SARS CoV 2 spike and angiotensin-converting enzyme 2

A recent experimental study found that the binding affinity between the cellular receptor human angiotensin converting enzyme 2 (ACE2) and receptor-binding domain (RBD) in spike (S) protein of novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is more than 10-fold higher than that of the original severe acute respiratory syndrome coronavirus (SARS-CoV). However, main-chain structures of the SARS-CoV-2 RBD are almost the same with that of the SARS-CoV RBD. Understanding physical mechanism responsible for the outstanding affinity between the SARS-CoV-2 S and ACE2 is the "urgent challenge" for developing blockers, vaccines and therapeutic antibodies against the coronavirus disease 2019 (COVID-19) pandemic. Considering the mechanisms of hydrophobic interaction, hydration shell, surface tension, and the shielding effect of water molecules, this study reveals a hydrophobic-interaction-based mechanism by means of which SARS-CoV-2 S and ACE2 bind together in an aqueous environment. The hydrophobic interaction between the SARS-CoV-2 S and ACE2 protein is found to be significantly greater than that between SARS-CoV S and ACE2. At the docking site, the hydrophobic portions of the hydrophilic side chains of SARS-CoV-2 S are found to be involved in the hydrophobic interaction between SARS-CoV-2 S and ACE2. We propose a method to design live attenuated viruses by mutating several key amino acid residues of the spike protein to decrease the hydrophobic surface areas at the docking site. Mutation of a small amount of residues can greatly reduce the hydrophobic binding of the coronavirus to the receptor, which may be significant reduce infectivity and transmissibility of the virus.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

DyHGCN: A Dynamic Heterogeneous Graph Convolutional Network to Learn Users' Dynamic Preferences for Information Diffusion Prediction

Information diffusion prediction is a fundamental task for understanding the information propagation process. It has wide applications in such as misinformation spreading prediction and malicious account detection. Previous works either concentrate on utilizing the context of a single diffusion sequence or using the social network among users for information diffusion prediction. However, the diffusion paths of different messages naturally constitute a dynamic diffusion graph. For one thing, previous works cannot jointly utilize both the social network and diffusion graph for prediction, which is insufficient to model the complexity of the diffusion process and results in unsatisfactory prediction performance. For another, they cannot learn users' dynamic preferences. Intuitively, users' preferences are changing as time goes on and users' personal preference determines whether the user will repost the information. Thus, it is beneficial to consider users' dynamic preferences in information diffusion prediction. In this paper, we propose a novel dynamic heterogeneous graph convolutional network (DyHGCN) to jointly learn the structural characteristics of the social graph and dynamic diffusion graph. Then, we encode the temporal information into the heterogeneous graph to learn the users' dynamic preferences. Finally, we apply multi-head attention to capture the context-dependency of the current diffusion path to facilitate the information diffusion prediction task. Experimental results show that DyHGCN significantly outperforms the state-of-the-art models on three public datasets, which shows the effectiveness of the proposed model.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

The role of hydrophobic interactions in folding of $β$-sheets

Exploring the protein-folding problem has been a long-standing challenge in molecular biology. Protein folding is highly dependent on folding of secondary structures as the way to pave a native folding pathway. Here, we demonstrate that a feature of a large hydrophobic surface area covering most side-chains on one side or the other side of adjacent $β$-strands of a $β$-sheet is prevail in almost all experimentally determined $β$-sheets, indicating that folding of $β$-sheets is most likely triggered by multistage hydrophobic interactions among neighbored side-chains of unfolded polypeptides, enable $β$-sheets fold reproducibly following explicit physical folding codes in aqueous environments. $β$-turns often contain five types of residues characterized with relatively small exposed hydrophobic proportions of their side-chains, that is explained as these residues can block hydrophobic effect among neighbored side-chains in sequence. Temperature dependence of the folding of $β$-sheet is thus attributed to temperature dependence of the strength of the hydrophobicity. The hydrophobic-effect-based mechanism responsible for $β$-sheets folding is verified by bioinformatics analyses of thousands of results available from experiments. The folding codes in amino acid sequence that dictate formation of a $β$-hairpin can be deciphered through evaluating hydrophobic interaction among side-chains of an unfolded polypeptide from a $β$-strand-like thermodynamic metastable state.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Topic Adaptation and Prototype Encoding for Few-Shot Visual Storytelling

Visual Storytelling~(VIST) is a task to tell a narrative story about a certain topic according to the given photo stream. The existing studies focus on designing complex models, which rely on a huge amount of human-annotated data. However, the annotation of VIST is extremely costly and many topics cannot be covered in the training dataset due to the long-tail topic distribution. In this paper, we focus on enhancing the generalization ability of the VIST model by considering the few-shot setting. Inspired by the way humans tell a story, we propose a topic adaptive storyteller to model the ability of inter-topic generalization. In practice, we apply the gradient-based meta-learning algorithm on multi-modal seq2seq models to endow the model the ability to adapt quickly from topic to topic. Besides, We further propose a prototype encoding structure to model the ability of intra-topic derivation. Specifically, we encode and restore the few training story text to serve as a reference to guide the generation at inference time. Experimental results show that topic adaptation and prototype encoding structure mutually bring benefit to the few-shot model on BLEU and METEOR metric. The further case study shows that the stories generated after few-shot adaptation are more relative and expressive.

0 citations0 reviewsNo code linkedOpen access