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

Fanxu Meng

Fanxu Meng contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate
Artificial IntelligenceMachine LearningInformation Theorymath.ITComputation and LanguageComputer VisionPerformancequant-ph

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
8works
0followers
8topics
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

8 published item(s)

preprint2026arXiv

BioPIE: A Biomedical Protocol Information Extraction Dataset for High-Reasoning-Complexity Experiment Question Answer

Question Answer (QA) systems for biomedical experiments facilitate cross-disciplinary communication, and serve as a foundation for downstream tasks, e.g., laboratory automation. High Information Density (HID) and Multi-Step Reasoning (MSR) pose unique challenges for biomedical experimental QA. While extracting structured knowledge, e.g., Knowledge Graphs (KGs), can substantially benefit biomedical experimental QA. Existing biomedical datasets focus on general or coarsegrained knowledge and thus fail to support the fine-grained experimental reasoning demanded by HID and MSR. To address this gap, we introduce Biomedical Protocol Information Extraction Dataset (BioPIE), a dataset that provides procedure-centric KGs of experimental entities, actions, and relations at a scale that supports reasoning over biomedical experiments across protocols. We evaluate information extraction methods on BioPIE, and implement a QA system that leverages BioPIE, showcasing performance gains on test, HID, and MSR question sets, showing that the structured experimental knowledge in BioPIE underpins both AI-assisted and more autonomous biomedical experimentation.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

GQLA: Group-Query Latent Attention for Hardware-Adaptive Large Language Model Decoding

Multi-head Latent Attention (MLA), the attention used in DeepSeek-V2/V3, jointly compresses keys and values into a low-rank latent and matches the H100 roofline almost perfectly. Its trained weights, however, expose only one decoding path - an absorbed MQA form - which ties efficient inference to H100-class compute-bandwidth ratios, forfeits tensor parallelism along the head axis, and yields no Multi-Token Prediction (MTP) gain on commodity inference GPUs such as the export-restricted H20. We propose Group-Query Latent Attention (GQLA), a minimal modification of MLA whose trained weights expose two algebraically equivalent decoding paths over the same parameters: an MQA-absorb path identical to MLA's, and a GQA path with a per-group expanded cache. The runtime picks the path that matches the target hardware - no retraining, no custom kernels - so a single set of GQLA weights pins the rooflines of both H100 (MQA-absorb, s_q=1) and H20 (GQA + MTP, s_q=2), while supporting up to 8-way zero-redundancy tensor parallelism on the GQA path. To avoid pretraining from scratch we extend TransMLA into TransGQLA, which converts a pretrained GQA checkpoint into a GQLA model; on LLaMA-3-8B it compresses the per-token KV cache to 28.125% of the GQA baseline on the MQA-absorb path while structurally preserving GQA-level traffic on the per-group path.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

MISA: Mixture of Indexer Sparse Attention for Long-Context LLM Inference

DeepSeek Sparse Attention (DSA) sets the state of the art for fine-grained inference-time sparse attention by introducing a learned token-wise indexer that scores every prefix token and selects the most relevant ones for the main attention. To remain expressive, the indexer uses many query heads (for example, 64 on DeepSeek-V3.2) that share the same selected token set; this multi-head design is precisely what makes the indexer the dominant cost on long contexts. We propose MISA (Mixture of Indexer Sparse Attention), a drop-in replacement for the DSA indexer that treats its indexer heads as a pool of mixture-of-experts. A lightweight router uses cheap block-level statistics to pick a query-dependent subset of only a few active heads, and only those heads run the heavy token-level scoring. This preserves the diversity of the original indexer pool while reducing the per-query cost from scoring every prefix token with every head to scoring it with only a handful of routed heads, plus a negligible router term computed on a small set of pooled keys. We further introduce a hierarchical variant of MISA that uses the routed pass to keep an enlarged candidate set and then re-ranks it with the original DSA indexer to recover the final selected tokens almost exactly. With only eight active heads and no additional training, MISA matches the dense DSA indexer on LongBench across DeepSeek-V3.2 and GLM-5 while running with eight and four times fewer indexer heads respectively, and outperforms HISA on average. It also preserves fully green Needle-in-a-Haystack heatmaps up to a 128K-token context and recovers more than 92% of the tokens selected by the DSA indexer per layer. Our TileLang kernel delivers roughly a 3.82 times speedup over DSA's original indexer kernel on a single NVIDIA H200 GPU.

0 citations0 reviewsNo code linkedOpen access
preprint2026arXiv

Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Models

We introduce Youtu-LLM, a lightweight yet powerful language model that harmonizes high computational efficiency with native agentic intelligence. Unlike typical small models that rely on distillation, Youtu-LLM (1.96B) is pre-trained from scratch to systematically cultivate reasoning and planning capabilities. The key technical advancements are as follows: (1) Compact Architecture with Long-Context Support: Built on a dense Multi-Latent Attention (MLA) architecture with a novel STEM-oriented vocabulary, Youtu-LLM supports a 128k context window. This design enables robust long-context reasoning and state tracking within a minimal memory footprint, making it ideal for long-horizon agent and reasoning tasks. (2) Principled "Commonsense-STEM-Agent" Curriculum: We curated a massive corpus of approximately 11T tokens and implemented a multi-stage training strategy. By progressively shifting the pre-training data distribution from general commonsense to complex STEM and agentic tasks, we ensure the model acquires deep cognitive abilities rather than superficial alignment. (3) Scalable Agentic Mid-training: Specifically for the agentic mid-training, we employ diverse data construction schemes to synthesize rich and varied trajectories across math, coding, and tool-use domains. This high-quality data enables the model to internalize planning and reflection behaviors effectively. Extensive evaluations show that Youtu-LLM sets a new state-of-the-art for sub-2B LLMs. On general benchmarks, it achieves competitive performance against larger models, while on agent-specific tasks, it significantly surpasses existing SOTA baselines, demonstrating that lightweight models can possess strong intrinsic agentic capabilities.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Filter Grafting for Deep Neural Networks: Reason, Method, and Cultivation

Filter is the key component in modern convolutional neural networks (CNNs). However, since CNNs are usually over-parameterized, a pre-trained network always contain some invalid (unimportant) filters. These filters have relatively small $l_{1}$ norm and contribute little to the output (\textbf{Reason}). While filter pruning removes these invalid filters for efficiency consideration, we tend to reactivate them to improve the representation capability of CNNs. In this paper, we introduce filter grafting (\textbf{Method}) to achieve this goal. The activation is processed by grafting external information (weights) into invalid filters. To better perform the grafting, we develop a novel criterion to measure the information of filters and an adaptive weighting strategy to balance the grafted information among networks. After the grafting operation, the network has fewer invalid filters compared with its initial state, enpowering the model with more representation capacity. Meanwhile, since grafting is operated reciprocally on all networks involved, we find that grafting may lose the information of valid filters when improving invalid filters. To gain a universal improvement on both valid and invalid filters, we compensate grafting with distillation (\textbf{Cultivation}) to overcome the drawback of grafting . Extensive experiments are performed on the classification and recognition tasks to show the superiority of our method. Code is available at \textcolor{black}{\emph{https://github.com/fxmeng/filter-grafting}}.

0 citations0 reviewsNo code linkedOpen access
preprint2021arXiv

Quantum Algorithm for DOA Estimation in Hybrid Massive MIMO

The direction of arrival (DOA) estimation in array signal processing is an important research area. The effectiveness of the direction of arrival greatly determines the performance of multi-input multi-output (MIMO) antenna systems. The multiple signal classification (MUSIC) algorithm, which is the most canonical and widely used subspace-based method, has a moderate estimation performance of DOA. However, in hybrid massive MIMO systems, the received signals at the antennas are not sent to the receiver directly, and spatial covariance matrix, which is essential in MUSIC algorithm, is thus unavailable. Therefore, the spatial covariance matrix reconstruction is required for the application of MUSIC in hybrid massive MIMO systems. In this article, we present a quantum algorithm for MUSIC-based DOA estimation in hybrid massive MIMO systems. Compared with the best-known classical algorithm, our quantum algorithm can achieve an exponential speedup on some parameters and a polynomial speedup on others under some mild conditions. In our scheme, we first present the quantum subroutine for the beam sweeping based spatial covariance matrix reconstruction, where we implement a quantum singular vector transition process to avoid extending the steering vectors matrix into the Hermitian form. Second, a variational quantum density matrix eigensolver (VQDME) is proposed for obtaining signal and noise subspaces, where we design a novel objective function in the form of the trace of density matrices product. Finally, a quantum labeling operation is proposed for the direction of arrival estimation of the signal.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Accurate Closed-Form Approximations to Channel Distributions of RIS-Aided Wireless Systems

This paper proposes highly accurate closed-form approximations to channel distributions of two different reconfigurable intelligent surface (RIS)-based wireless system setups, namely, dual-hop RIS-aided (RIS-DH) scheme and RIS-aided transmit (RIS-T) scheme. Differently from previous works, the proposed approximations reveal to be very tight for arbitrary number $N$ of reflecting metasurface&#39;s elements. Our findings are then applied to the performance analysis of the considered systems, in which the outage probability, bit error rate, and average channel capacity are derived. Results show that the achievable diversity orders $G_d$ for RIS-DH and RIS-T schemes are $N-1<G_d<N$ and $N$, respectively. Furthermore, it is revealed that both schemes can not provide the multiplexing gain and only diversity gains are achieved. For the RIS-DH scheme, the channels are similar to the keyhole multiple-input multiple-output (MIMO) channels with only one degree of freedom, while the RIS-T scheme is like the transmit diversity structure.

0 citations0 reviewsNo code linkedOpen access
preprint2020arXiv

Filter Grafting for Deep Neural Networks

This paper proposes a new learning paradigm called filter grafting, which aims to improve the representation capability of Deep Neural Networks (DNNs). The motivation is that DNNs have unimportant (invalid) filters (e.g., l1 norm close to 0). These filters limit the potential of DNNs since they are identified as having little effect on the network. While filter pruning removes these invalid filters for efficiency consideration, filter grafting re-activates them from an accuracy boosting perspective. The activation is processed by grafting external information (weights) into invalid filters. To better perform the grafting process, we develop an entropy-based criterion to measure the information of filters and an adaptive weighting strategy for balancing the grafted information among networks. After the grafting operation, the network has very few invalid filters compared with its untouched state, enpowering the model with more representation capacity. We also perform extensive experiments on the classification and recognition tasks to show the superiority of our method. For example, the grafted MobileNetV2 outperforms the non-grafted MobileNetV2 by about 7 percent on CIFAR-100 dataset. Code is available at https://github.com/fxmeng/filter-grafting.git.

0 citations0 reviewsNo code linkedOpen access