跳到正文
研踪PaperTrail

版本更新

查看已发布论文的最新修订,在详情页对照实际观测到的版本。

部分采集异常

全站更新概况

+1085新增论文
+884其中 AI 相关
+130已有论文修订
+94新增 AI 导读

arXiv 最近完成扫描:2026-11-19 06:59 · 全站最近入库:2026-11-19 06:59各信源进度 →

统计口径与数据新鲜度

统计 2026-11-18 08:21 — 2026-11-19 08:21(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-19 06:59最近发现新增版本:2026-11-19 06:59库内论文最新更新时间:2026-11-18 08:00

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

26,871 已收录论文22,375 AI 相关8,156 有修订112 HF 社区发现

论文修订列表

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 8,156 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-19 06:57(北京时间)
版本更新 · v2arXiv:2611.00749

PeptSol: A Translational Benchmark for Protected-Peptide Solubility in Drug Manufacturing

Alexander Arjun Lobo, Marc Saouda, Urvi Awasthi, Leonid Zhukov

原文摘要

Liquid-phase peptide synthesis (LPPS) offers a scalable route to peptide manufacturing, but route development still depends on empirical solvent screens because each protected intermediate must be soluble for reaction and recoverable for purification. The protecting groups, terminal modifications, and noncanonical residues that govern this behaviour are poorly represented in existing peptide property benchmarks, limiting the ability to develop and rigorously evaluate molecular representation models for LPPS-relevant prediction tasks. We introduce PeptSol, an open benchmark of 1388 hand-curated protected-peptide solubility measurements with chemistry-aware splits. The measurements span only 84 parent sequences and 101 protected structures, and 92.5% come from a single historical programme. Across seven frozen molecular encoders, the best macro-F1 is 0.843 on binary solubility for the historical subset under a peptide-held-out split; the harder three-class ordinal task reaches 0.614 in that cell. No foundation model consistently outperforms a baseline combining chemical fingerprints with explicit molecular descriptors, and model rankings vary across prediction tasks and data splits. The dataset, evaluation protocol, and baselines establish a standard for measuring progress on molecular representations for synthesis-aware peptide developability.

cs.LG · physics.chem-ph · q-bio.BM

版本更新 · v2arXiv:2610.33258

Two Uncertainties Are Not One: Disagreement versus Decision-Risk in Cheap Structured Content Moderation

Karan Yadav

原文摘要

Content moderation compresses a graded, contested human judgment into a binary act (leave up / take down) over billions of items, at a scale where running a frontier LLM on every post is uneconomical. A new class of frozen "System One" models returns typed, calibrated decisions instead of text at ~USD 0.000025 per call and ~100 ms. We run the first independent audit of one such model on 142,658 human-labeled comments, and benchmark it against 14 frontier reasoning models (the GPT-5.x, Claude, Gemini 3, and Grok families) on a shared probe set. Two findings stand out. First, the cheap model is accuracy-competitive: its zero-shot classification AUROC (0.83) is not statistically separable (paired bootstrap, n=56; a failure to reject parity, not proof of equivalence) from 13 of the 14 frontier systems while costing 1-3 orders of magnitude less, and it is no worse calibrated than several of them. Second, "confidence" conflates two uncertainties that come apart: human disagreement (where annotators split on satire and coded language) and decision risk (where the automatic action is costly-wrong). Under an asymmetric moderation loss these call for opposite escalation signals, so no single confidence number serves both. Because the model prices extra questions at nearly zero, decomposing the judgment into parallel sub-questions and reading their spread recovers a human-disagreement signal that the model's native confidence misses -- but this remedy is specific: it helps the weakest, disagreement-blind models and only when the decomposition aligns with the sources of disagreement; a specialist toxicity model and the strongest LLMs do not need it, and the effect is distribution-sensitive. We will release the instrument, the cached runs, and the analysis.

cs.CL · cs.CE · cs.LO

版本更新 · v2arXiv:2606.01483

Murmur: Efficient Long-Context ASR Inference with Hierarchical Chunking

Wei-Tzu Lee, Keisuke Kamahori, Baris Kasikci

原文摘要

Long-form automatic speech recognition (ASR) requires both high accuracy and low inference latency, but existing systems fall short of handling the trade-off between the two. Chunk-based approaches process audio in parallel windows for low latency, but lose cross-chunk context and need brittle heuristics to align speakers and timestamps at boundaries. Long-context ASR models transcribe everything in a single pass for better accuracy, but run several times slower, and a single repetition loop can corrupt the entire output. In this paper, we first analyze how context length affects the accuracy of long-context ASR models and find that an intermediate chunk size of about five minutes matches single-pass accuracy while avoiding repetition loops. Building on this, we propose Murmur, an inference system for long-form ASR based on hierarchical chunking. The long-context model decodes intermediate target chunks in parallel, and a fast short-window ASR on 30 s draft windows accelerates each chunk with lossless, structure-aware speculative decoding. The short-window ASR drafts the transcript, simple format rules draft the timestamps and speaker tags, and the long-context model verifies every token. On two long-context ASR models and four meeting datasets in English and Mandarin, Murmur cuts end-to-end latency per recording by 2.5x on average (1.9-2.9x) relative to single-pass decoding, including diarization, while lowering tcpWER by 1.9 points on average. Our code is available at https://github.com/uw-syfi/Murmur.

cs.LG · cs.AI · eess.AS

版本更新 · v2arXiv:2607.13033

DenseReward: Dense Reward Learning via Failure Synthesis for Robotic Manipulation

Yu Fang, Wanxi Dong, Jiaqi Liu, Yue Yang, Mingxiao Huo, Yao Mu, Huaxiu Yao, Li Erran Li 等 10 位作者

原文摘要

Reinforcement learning holds great promise for improving robot policies beyond the limits of imitation learning. However, its practical adoption remains bottlenecked by the lack of reliable vision-language reward models that provide dense and informative feedback. Two key challenges remain: acquiring diverse failure data at scale and obtaining fine-grained reward signals beyond sparse trajectory-level success labels. Existing approaches reuse successful trajectories to construct negative or suboptimal examples through trajectory truncation, instruction relabeling, or trajectory comparisons, without generating new trajectories where physical interaction failures such as collisions, missed grasps, object drops, and recovery actually occur. Meanwhile, many existing reward models predict sparse binary or trajectory-level rewards, which provide limited guidance for efficient policy optimization. We introduce DenseReward, a dense robotic reward model that addresses both challenges. To train DenseReward, we develop an automated failure data generation pipeline that synthesizes physically realistic failure trajectories in simulation without manual annotation, covering diverse failure modes such as collisions, missed grasps, object drops, and recovery behaviors. DenseReward predicts dense frame-level reward scores from visual observations and language instructions, enabling fine-grained estimation of task progress throughout an episode. Experiments show that DenseReward outperforms general-purpose VLMs and existing robotic reward models in dense reward prediction across both simulated and real-world manipulation. We further demonstrate that DenseReward provides effective reward guidance for downstream model predictive control and reinforcement learning.

cs.RO

版本更新 · v6arXiv:2601.22478

Transformation-Augmented GRPO for Enhancing Exploration in Reasoning of Large Language Models

Khiem Le, Phuc Nguyen, Youssef Mroueh, Chi-Heng Lin, Shangqian Gao, Ting Hua, Nitesh V. Chawla

原文摘要

Group Relative Policy Optimization (GRPO) has become the dominant reinforcement learning method for eliciting reasoning capabilities in LLMs using verifiable rewards. Still, it suffers from two critical issues that limit models' exploration: \textbf{Gradient vanishing} and \textbf{Diversity collapse}. When a training question is too easy or too hard, all sampled responses receive identical rewards, yielding no gradient signal. Meanwhile, the model tends to collapse its responses toward a single reasoning pattern rather than exploring diverse strategies. We propose Transformation-Augmented GRPO (TA-GRPO), a simple yet effective method that addresses both issues via question transformation. For each training question, we automatically create multiple problem-equivalent rephrased questions that alter wording, format, and information order while preserving the underlying intent. As these rephrasings shift the model's perceived difficulty, responses across the original question and its rephrasings receive mixed rewards and encode more diverse reasoning strategies within it. TA-GRPO jointly computes advantages over this expanded response group and aligns all importance sampling ratios to the original question, enabling the model to learn from a richer set of attempts. Experiments with four LLMs (Qwen3-1.7B/4B-Base and Llama-3.2-1B/3B-Instruct) show that TA-GRPO consistently improves pass@$k$ on competition-level mathematics benchmarks (AMC, OlympiadBench, AIME24, AIME25) and out-of-distribution benchmarks (Minerva, GPQA-Diamond). Notably, it improves the average pass@32 of Qwen3-1.7B-Base and Qwen3-4B-Base by \textbf{4.97}\% and \textbf{4.35}\%, respectively, and matches the pass@32 of baselines trained with up to 2.5x more data.

cs.LG

版本更新 · v2arXiv:2604.13699

MIND: AI Co-Scientist for Material Research

Geonhee Ahn, Donghyun Lee, Hayoung Doo, Jonggeol Na, Hyunsoo Cho, Sookyung Kim

原文摘要

Large language models (LLMs) have enabled agentic AI systems for scientific discovery, but most approaches remain limited to textbased reasoning without automated experimental verification. We propose MIND, an LLM-driven framework for automated hypothesis validation in materials research. MIND organizes the scientific discovery process into hypothesis refinement, experimentation, and debate-based validation within a multi-agent pipeline. For experimental verification, the system integrates Machine Learning Interatomic Potentials, particularly SevenNet-Omni, enabling scalable in-silico experiments. We also provide a web-based user interface for automated hypothesis testing. The modular design allows additional experimental modules to be integrated, making the framework adaptable to broader scientific workflows. The code is available at: https://github.com/IMMS-Ewha/MIND, and a demonstration video at: https://youtu.be/lqiFe1OQzN4.

cs.MA · cs.AI · cs.CE

版本更新 · v2arXiv:2603.13665

An Extended Study of Gear-Ratio-Aware Standard Cell Layout Generation for DTCO Exploration

Chung-Kuan Cheng, Andrew B. Kahng, Bill Lin, Yucheng Wang, Dooseok Yoon

原文摘要

Advanced nodes decouple contacted poly pitch (CPP) and lower-metal pitch to improve routability. We present CPCell, an efficient standard-cell layout generation framework, to support arbitrary gear ratio (GR) and offset parameters through a fine-grained layered grid graph and constraint-programming-based placement-routing co-optimization. Layout quality is improved via Middle-of-Line routing, M0 pin enablement, pin accessibility constraints and a weighted multi-objective formulation that jointly optimizes cell layouts. To scale to netlists with up to 48 transistors, we incorporate acceleration techniques including transistor clustering, identical transistor partitioning, routing lower bound tightening and early termination strategies. Comprehensive cell-level and block-level studies are conducted to evaluate GR and offset choices, quantify the benefits of the proposed objectives and assess their impact on power, performance, area and IR-drop outcomes.

cs.AR

版本更新 · v2arXiv:2610.03056

MOF-VERIFY: A Failure-Aware Agentic Harness for MOF Hypothesis Verification

Donghyun Lee, Taehoon Lee, Geonhee Ahn, Jieun Kim, Jihyun Park, Suyeon Cho, Yoona Kim, Chaerim Shin 等 13 位作者

原文摘要

Large language models are increasingly used as reasoning components in AI-driven materials Co-Scientists, yet the reliability of the resulting verification pipeline remains unclear. Metal-organic frameworks (MOFs) provide a particularly challenging setting because structures may appear under different identifiers, synthesis outcomes depend strongly on experimental conditions, evidence is distributed across heterogeneous sources, and some hypotheses require computation rather than literature alone. We introduce a diagnostic benchmark with four task families covering structural grounding, synthesis-condition verification, evidence-sufficiency verification, and MLIP-based computational verification. T-MOF-1-3 are evaluated under closed-book, retrieval-enabled, and oracle-evidence settings to localize failures in knowledge access, evidence acquisition, and reasoning, while T-MOF-4 separately evaluates computational verification. Guided by these diagnosed failure modes, we develop MOF-Verify, a failure-aware agentic harness that targets structural, literature, evidence-sufficiency, and computational bottlenecks before producing a final verdict. Across multiple backbone LLMs, MOF-Verify improves the reported average over direct inference and retrieval-based baselines on this benchmark, with gains concentrated in structural grounding and evidence sufficiency. Benchmark records and release-status documentation are available at https://github.com/IMMS-Ewha/MOF-Verify-Benchmark.

cs.AI · cs.CE

版本更新 · v2arXiv:2605.15391

PanoWorld: Geometry-Consistent Panoramic Video World Modeling

Le Jiang, Xiangyu Bai, Bishoy Galoaa, Shayda Moezzi, Caleb James Lee, Tooba Imtiaz, Edmund Yeh, Jennifer Dy 等 10 位作者

原文摘要

We present PanoWorld, a panoramic video world model that generates geometry-consistent 360$\degree$ video from a single image and a caption. Existing panoramic video methods optimize primarily for visual realism and do not explicitly constrain the underlying 3D scene state, producing outputs that appear plausible yet exhibit inconsistent depth, broken correspondences, and implausible motion across the spherical surface. We address this gap by framing panoramic video generation as a geometry- and dynamics-consistent latent state modeling problem rather than pure visual synthesis. Building on a pre-trained perspective video world model, we introduce two lightweight regularizers: a depth consistency loss against pseudo ground-truth panoramic depth, and a trajectory consistency loss that supervises the 3D world-frame positions of tracked points across time. We further apply spherical-geometry-aware adaptation to the conditioning and positional encoding. We additionally introduce PanoGeo, a unified geometry-aware panoramic video dataset with consistent depth, trajectory, and prompt annotations across diverse real and synthetic sources, used for both training and stratified evaluation. Experiments show that PanoWorld improves geometric consistency over prior panoramic generation methods while maintaining competitive visual realism, establishing that panoramic video generation must be treated as a geometric modeling problem to support the holistic spatial understanding requirements of embodied AI applications. Code is available at https://github.com/ostadabbas/PanoWorld.

cs.CV · cs.AI

版本更新 · v2arXiv:2605.06458

Invariant Features in Language Models: Geometric Characterization and Model Attribution

Agnibh Dasgupta, Abdullah Tanvir, Xin Zhong

原文摘要

Language models exhibit strong robustness to paraphrasing, suggesting that semantic information may be encoded through stable internal representations, yet the structure and origin of such invariance remain unclear. We propose a local geometric framework in which semantically equivalent inputs produce representations that occupy structured regions in latent space, with paraphrastic variation along nuisance directions and semantic identity preserved in invariant subspaces. Building on this view, we make three contributions: (1) a geometric characterization of invariant latent features, (2) a contrastive subspace discovery method that separates semantic-changing from semantic-preserving variation, and (3) an application of invariant representations to zero-shot model attribution. Across models and layers, empirical results support these contributions. Invariant structure emerges in specific depth regions, semantic displacement lies largely outside the nuisance subspace, and representation-level interventions indicate a causal role of invariant components in model outputs. Invariant representations also capture model-specific geometric patterns, enabling accurate attribution. Together, these findings support a local geometric characterization of semantic invariance in language model representations.

cs.LG · cs.CL

版本更新 · v3arXiv:2011.04672

Principles of Quantum Communication Theory: A Modern Approach

Sumeet Khatri, Ludovico Lami, Mark M. Wilde

原文摘要

This is a preliminary version of a book in progress on the theory of quantum communication. We adopt an information-theoretic perspective throughout and give a comprehensive account of fundamental results in quantum communication theory from the past decade (and earlier), with an emphasis on the modern one-shot-to-asymptotic approach that underlies much of today's state-of-the-art research in this field. In Part I, we cover mathematical preliminaries and provide a detailed study of quantum mechanics from an information-theoretic perspective. We also provide an extensive and thorough review of quantum entropies, and we devote an entire chapter to the study of entanglement measures. Equipped with these essential tools, in Part II we study classical communication (with and without entanglement assistance), entanglement distillation, quantum communication, secret key distillation, and private communication. In Part III, we cover the latest developments in feedback-assisted communication tasks, such as quantum and classical feedback-assisted communication, LOCC-assisted quantum communication, and secret key agreement.

quant-ph · cond-mat.stat-mech · cs.IT · hep-th · math-ph

版本更新 · v3arXiv:2608.04463

The Judge is Part of the Experiment: Auditing Peer Context Sensitivity in LLM-Based Evaluation

Alicia Guerra, Jiaming Qu, Yibo Hu

原文摘要

Large language model (LLM)-based judges are increasingly used to evaluate AI systems. An underexplored area is applying these judges to multi-agent systems, where they may see the interactions behind a final answer. These interactions, which we call peer context, can affect an evaluation in two ways: (1) they can change the answer produced by the multi-agent system, because an agent might be swayed by other agents, and (2) they can change how a judge evaluates that same answer, depending on whether the judge sees them. In this paper, we separate these two effects through a controlled experiment. We first let an agent answer a question and then answer again after seeing peer responses. We then prompt judges to rate the same final answer with and without the peer context. Finally, we check whether the judges' ratings establish a valid measure of answer quality. Across three benchmarks, four open-weight LLMs as agents, and six judges, we find that peer context changes not only the answer but also how an identical answer is judged, and judges react to it in opposite directions. Moreover, providing peer context for a wrong answer misleads agents more than correct peers help them; without the correct answer, judges struggle to tell known-correct from known-wrong answers---a problem that a well-fitted statistical model does not reveal. Our results show that even when evaluating the same answer, how a judge is set up can change the conclusions differently. We therefore call for evaluations to report what each judge sees, document detailed results per judge, and validate rating scales. These practices are an important step toward the accountable use of LLM-based judges.

cs.CL

版本更新 · v3arXiv:2606.05647

Coding with the Enemy: Can Human Developers Detect AI Agent Sabotage?

Jingheng Ye, Huiqi Zou, Simon Yu, Weiyan Shi

原文摘要

AI coding agents are increasingly embedded in real-world software development, collaborating with human developers while gaining broader access to codebases and tools. This creates a new attack surface: an agent can exploit human trust to sabotage development, for instance by inserting malicious code to accomplish a hidden side task. Most prior work studies AI sabotage in AI-only settings, paying limited attention to the role of human oversight in detecting and mitigating such malicious behavior. To address this gap, we conduct the first large-scale study of human oversight in AI coding sabotage. Over 100 participants collaborate with one of four frontier models (Claude-Opus-4.6, GPT-5.4, Gemini-3.1-Pro, and MiniMax-M2.7) on a long-horizon coding task lasting around five hours, designed to mimic real-world workflows. We find that 83/88 (94%) of developers in the no-monitor conditions fail to detect sabotage, and our analysis of participant feedback attributes this vulnerability to minimal code review, plausible cover story, and overtrust in agents. We further test the effectiveness of a safety monitor in one condition: while the monitor reduces sabotage success, sabotage still succeeds in 9/16 (56%) of sessions with a correct monitor alert. Drawing on participant feedback, we offer actionable suggestions for better monitor design. This work complements existing AI safety research and highlights an urgent need for human-centric safety mechanisms that account for human factors, particularly in long-horizon, real-world development settings.

cs.AI · cs.CL · cs.CY · cs.HC

版本更新 · v2arXiv:2601.00878

Heraclitean Dialectical Concept Space

Volkan Yildiz

原文摘要

We introduce Heraclitean Dialectical Concept Spaces (HDCS), a topological framework for modelling stagewise structural change in systems of concepts. A stage consists of a nonempty concept set, a feasible family generating an external topology, and a neighbourhood assignment carrying additional relational structure. Neighbourhood overlap induces adjacency and channel ideals. Candidate emergence is represented by the CCER construction: an interaction profile, finite admissible cores, and anchored witnesses determine nonempty open candidate regions in uncovered remainder structure. Selected candidates may be reified as new atomic concepts at a later stage, while partial injective carry maps separately track persistence of existing identities. The resulting history space is obtained as a quotient of the disjoint union of stage spaces with the corresponding final topology. Product concept spaces extend the construction to cross-domain interaction. We derive support characterizations, mediated-interaction components, canonical CCER envelopes, monotonicity and impossibility criteria, product-space factorization, and invariance under concept-space isomorphism. Applications to monetary transformation, zero and representation, mammalian middle-ear evolution, and AI-related labour change illustrate the framework. HDCS is intended as a structural formalism rather than a causal or predictive theory of historical change.

physics.soc-ph · cs.LO · math.GN

版本更新 · v2arXiv:2605.30621

Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents

Minhua Lin, Juncheng Wu, Zijun Wang, Zhan Shi, Yisi Sang, Bing He, Zewen Liu, Tianxin Wei 等 17 位作者

原文摘要

LLM agents are increasingly deployed as systems built around editable external harnesses, including prompts, skills, memories and tools, that shape task execution without changing model parameters. Harness self-evolution adapts such agents by updating these harnesses from execution evidence. Yet it remains unclear whether a model's base capability in task-solving predicts its capabilities in harness self-evolution: which models produce useful harness updates, and which actually benefit from them? We analyze two harness self-evolution capabilities: (i) harness-updating, the capability to produce useful persistent harness updates from execution evidence; (ii) harness-benefit, the capability to benefit from updated harnesses during task solving. Our analysis reveals two findings. First, harness-updating is flat in base capability: models from different capability tiers produce harness updates that lead to surprisingly similar gains; even Qwen3.5-9B's updates yield gains comparable to those of Claude Opus~4.6. Second, harness-benefit is non-monotonic in base capability: weak-tier models benefit little from updated harnesses, mid-tier models benefit most, and strong-tier models benefit less than mid-tier. We trace low gains at the weak tier to two failure modes: weak-tier models may fail to activate relevant harness artifacts, or activate them but fail to follow them faithfully. These findings suggest investing capability budget in the task-solving agent rather than the evolver, and targeting harness invocation and long-horizon instruction following in agent training. Our source code is publicly available at https://github.com/A-EVO-Lab/a-evolve/tree/release/harness-evolution.

cs.AI

版本更新 · v4arXiv:2503.13726

Towards Energy- and QoS-aware Load Balancing for 5G Advanced: Leveraging O-RAN to Achieve Sustainability and Energy Efficiency

Gustavo Z. Bruno, Gabriel M. Almeida, Aloizio Da Silva, Luiz A. DaSilva, Joao F. Santos, Alexandre Huff, Kleber V. Cardoso, Cristiano B. Both

原文摘要

The increasing energy consumption of next-generation mobile networks necessitates the adoption of autonomous and energy-aware management strategies. This article proposes a novel adaptive solution leveraging the Open Radio Access Network (O-RAN) architecture to optimize energy efficiency while maintaining Quality of Service (QoS). The proposed approach introduces a hierarchical O-RAN-aligned control framework in which a Non-Real-Time RAN Intelligent Controller (Non-RT RIC) periodically computes energy-aware policies from long-term historical data, while a Near-Real-Time RIC enforces those policies through per-User Equipment (UE) handover actions on a sub-second timescale. We formulate a joint energy- and QoS-aware load balancing problem as a Mixed-Integer Non-Linear Programming (MINLP) model. This model optimizes UE association across E2 Nodes to minimize transmission power and autonomously deactivate underutilized cells, while satisfying strict QoS constraints. To validate the proposed solution, we deploy it in an experimental environment that simulates massive sports events. Experimental results demonstrate 72% energy savings over a 24-hour trace and sub-second handover delays (0.123-0.204 s/UE), confirming the solution's feasibility for autonomous and sustainable 5G Advanced networks. This work provides guidelines for deploying energy-efficient strategies in O-RAN environments and underscores the potential of adaptive solutions for sustainable mobile communications.

cs.NI

版本更新 · v2arXiv:2610.00663

Backdoor Containment via Expert Quarantine and Shutdown in LLMs

Jianwei Li, Min-Seon Kim, Jung-Eun Kim

原文摘要

Backdoored large language models (LLMs) can behave normally on benign inputs while producing attacker-specified outputs under hidden triggers. Existing defenses span four stages---prior-training, in-training, post-training, and inference-time---and share one of two underlying strategies: either suppress backdoor learning (by filtering poisoned data or interrupting its acquisition during optimization) or learn, then purify (by repairing model weights or gating inputs after a fully backdoored model has formed). We propose a third strategy, learn, but channel: allow backdoor formation during training but route it into a designated, quarantined component that can be disabled at deployment. To this end, we propose Quarantined Expert Shutdown QES, a computationally efficient containment strategy built in a regularization-steered MoE-like setting. Specifically, given a poisoned dataset, QES augments a Transformer-based language model with routed expert-specific LoRA branches and lightweight routers, and uses auxiliary routing objectives to attract trigger-conditioned behavior into a designated expert while preserving benign capability elsewhere. At deployment, mitigation reduces to a single constant-time operation: zeroing the quarantined expert's routing weight, without trigger screening or further updating model weights. Empirically, our methods reduce the attack success rate (ASR) from 100% to 0-10% on most settings across two tasks, three attacks, and four model families, while downstream utility is often preserved or only modestly affected. These results establish learn, but channel as a previously unexplored regime for backdoor containment in generative LLMs.

cs.AI · cs.CR · cs.LG

版本更新 · v2arXiv:2611.01335

Profile-aware diagnostics of discarded parameter-Hessian mass

Maxim Bolshim, Alexander Kugaevskikh

原文摘要

Evaluating block-diagonal curvature models involves two questions: how much curvature they omit, and what a specific correction changes at a finite step. We measure the squared off-block fraction of the parameter Hessian against references for group sizes and the observed within-group curvature profile. In CIFAR-100 ResNet-18 and a skip-free counterpart, similar initial discarded fractions accompany different profiles. At saved trained ResNet-18 weights, changing from stage to layer groups moves estimated off-block mass from below to above the profile envelope; an approximate conditional screen resolves this comparison in all six seed/input cells, while some skip free comparisons remain unresolved. Separately, in saved ResNet-50/Stanford Cars states, we compare candidates formed from off-group Hessian action, the remaining model residual and their combination through the same Hessian-informed map. At unit scale off-group input alone raises native minibatch loss relative to the base preconditioner candidate in 17 of 18 state/geometry comparisons, yet the combined candidate has lower loss than the remainder-only candidate in nine of those cases; preferences vary with scale and correction-norm matching. The profiles help assess block models and alternative groupings, while same-state candidate comparisons test a particular correction without equating structural mass with training gain.

cs.LG

版本更新 · v3arXiv:2609.22818

The Price of Safety: Benign-Case Utility and Token Overhead of Memory-Poisoning Defenses in LLM Agents

Pritom Bhowmik

原文摘要

Memory-poisoning defenses for LLM agents are usually judged by how well they stop attacks, yet almost all the traffic an agent sees is benign. We measure what these defenses cost when no attack is present. Our harness holds the backbone model, memory backend, retrieval, answer prompt, and judge fixed and changes only the defense. Both LLM-based defenses run on the agent's own backbone, and an integrity check verifies every run. We test four defenses on half of LoCoMo: five long conversations (750 turns and 255 questions per run), with each condition run three times. Sanitization and provenance checking never fired. LLM-based anomaly detection blocked 9 of 2,250 turns (0.4%). A read-time reranker dropped 2.2% of the memories it judged, withholding at least one memory from 14.9% of questions. A blind audit found content-based justification for only 2 of 50 sampled drops. No defense showed a detectable accuracy cost, at 95% intervals of $\pm$2.4 to $\pm$4.6 points; the reranker changed accuracy by +0.6 points (95% CI [-2.6, +3.8]). Anomaly detection and the stacked condition gained 4.7 points, but a blocked turn changes the memories extracted after it, so we do not count this as a benefit. Token overhead was 0% for sanitization and provenance, 1.5% for anomaly detection, 3.0% for the reranker, and 4.5% with all four stacked. An earlier version of this study ran the reranker condition on a different model than its baseline. A baseline on that same model shows the swap produced 2.9 of the 4.4 points of accuracy cost we then reported. Attack success rate reveals none of this, so we report benign cost alongside it. We release the harness, labels, and per-run outputs.

cs.CR · cs.AI

版本更新 · v2arXiv:2609.03079

LeanStream: A Speculate-and-Refine Streaming Framework for Efficient on-Device LLM Inference

Renyuan Liu, Yuyang Leng, Kaiyan Liu, Yuzhou Zhong, Shaohan Hu, Chun-Fu, Chen, Peijun Zhao 等 10 位作者

原文摘要

On-device LLM inference is attractive for privacy and responsiveness, but remains challenging on mobile and embedded devices because model weights far exceed available DRAM. Prior systems exploit activation sparsity and offload weights to SSD or flash storage, but face a fundamental systems trade-off: accurate sparse execution decisions require the latest context, whereas efficient computation-I/O overlap requires early prediction. As a result, existing designs either serialize execution or incur redundant weight fetches, extra computation, and large cache overheads. We present LeanStream, a streaming speculate-and-refine framework for efficient on-device LLM inference. LeanStream progressively refines computation, loading, and cache-retention priorities using partial GPU results, enabling fine-grained overlap between GPU execution and storage I/O. We implement LeanStream on both mobile and embedded platforms. Compared with prior on-device LLM inference systems, LeanStream reduces memory usage by 4.8$\times$ to 7.5$\times$ at the best throughput achieved by prior work, while further improving token generation throughput by 1.6$\times$ to 2.1$\times$. Source code is available at https://github.com/RenyuanLIU/LeanStream.

cs.LG

版本更新 · v2arXiv:2609.23947

Optimal analysis of Greedy for stochastic online Euclidean matching: flat torus to the rescue

Mingwei Yang, Sophie H. Yu

原文摘要

We study online metric matching with $n$ servers and $n$ requests, all sampled independently and uniformly from $[0,1]^d$. We analyze the Greedy algorithm, which matches each arriving request to its closest available server. Despite its strong empirical performance, non-trivial bounds on Greedy's competitive ratio were previously known only for $d = 1$, where it is $O(1)$-competitive [BFP23]. We prove that Greedy is $O(1)$-competitive for every fixed $d\geq3$ and has competitive ratio $Θ(\sqrt{\log n})$ for $d=2$. Our proof first analyzes Greedy on the flat torus and then transfers the resulting estimates to the cube.

cs.DS

版本更新 · v2arXiv:2610.20331

Are We Overthinking One-Step Generation?

Woobin Park, Jaeyeon Kim, Yilun Du, Sitan Chen, Bohyung Han

原文摘要

Diffusion and flow matching are now dominant frameworks for generative modeling in continuous domains, but they typically require multiple network evaluations at inference. This has motivated extensive research on one-step generative modeling primarily built on diffusion and flow, though existing approaches often rely on complex training machinery. In this work, we introduce Priming Models, a simple two-stage framework for training one-step generators. Motivated by the difficulty of distributional training from scratch, Priming Models are first trained on a reconstruction objective, e.g., pixel-space denoising at a fixed noise level, before undergoing feature-space distributional post-training. Our central scientific finding is that this simple reconstruction stage is already enough to entirely alleviate the difficulty of distributional training. Empirically, across class-conditional image generation, text-to-image generation, and world modeling, Priming Models significantly advance the Pareto frontier of training compute vs. generation quality. On ImageNet at $256 \times 256$ resolution, we achieve a state-of-the-art FID of $1.19$ among one-step generators using approximately one-tenth the training compute of prior methods with comparable FID. We open-source our codebase at https://github.com/JaeyeonKim01/priming.

cs.LG

版本更新 · v2arXiv:2610.38210

GAAC: Lossless KV Compression for Bandwidth-Efficient LLM Inference on Personal Devices

Yipin Guo, Arun M George, Siddharth Joshi

原文摘要

Weight-only, lossy quantization substantially reduces the storage and memory requirements for serving large language models (LLMs), but the growing key-value (KV) cache remains a bottleneck for long-context inference on personal devices. We present Group-Adaptive Additive Codec (GAAC), a lossless KV compression and serving system that exploits exponent redundancy to reduce storage and accelerate attention without introducing KV approximation. GAAC encodes BF16 exponents using calibrated per-channel base-plus-offset codes over contiguous ranges. GAAC adapts code widths across small groups while preserving KV values bit-for-bit. A fused attention kernel reconstructs KV tiles on the fly, avoiding full-cache decompression. Block-wise compression and fixed-capacity slots support incremental cache updates and paged serving. Across five tested models, GAAC achieves 1.405-1.407$\times$ lossless BF16 KV compression. We also study compression performance on low-precision tensors FP8 E4M3 caches ( 1.09-1.14$\times$) and E5M2 caches (1.25-1.30$\times$). On NVIDIA DGX Spark, GAAC accelerates decode attention at a 64K-token context by 1.19-1.27$\times$ and end-to-end W4A16 decode steps by 1.04-1.15$\times$. In serving, it reduces resident KV memory by approximately 26% and increases aggregate throughput by 13-29% under comparable KV-memory budgets. On Apple M4, GAAC achieves 1.02-1.25$\times$ attention speedups, demonstrating portability beyond NVIDIA GPUs. Code is available at https://github.com/Intelligent-Microsystems-Lab/GAAC.

cs.AI

版本更新 · v3arXiv:2503.13344

DeepSTEP: Simultaneous Tracking and Estimation of Pose for Animals and Humans

Shashikant Verma, Harish Katti, Soumyaratna Debnath, Yamuna Swamy, Shanmuganathan Raman

原文摘要

Maintaining a consistent identity across video frames while estimating pose is crucial for downstream tasks such as behavior analysis and action recognition, yet existing pipelines largely decouple tracking and pose estimation. Top-down methods rely on per-frame detections followed by identity association, leading to increased computational cost and error propagation, whereas bottom-up approaches often struggle to preserve identity reliably in multi-instance settings. To address these limitations, we introduce DeepSTEP (STEP), a Transformer-based framework for joint tracking and pose estimation across diverse animal species and humans. STEP unifies localization, identity maintenance, and keypoint estimation into a single model, eliminating the need for per-frame detections or post hoc matching. Our approach leverages discriminative model weight prediction to propagate target identity across frames while predicting anatomically meaningful keypoints. Discriminative model prediction typically requires predefined target states. In contrast, STEP assumes only an initial bounding box in the first frame, removing the need for explicit keypoint initialization. This is enabled by two components: Gaussian Map Soft Prediction (GMSP) and Offset Map Regression Adapter (OMRA), which allow direct keypoint prediction and simplify inference. Starting from the initial bounding box, STEP tracks the subject and estimates its pose in subsequent frames. We train and evaluate STEP on datasets spanning multiple species, demonstrating strong performance and greater efficiency than existing methods. By jointly modeling tracking and pose, STEP provides an effective and scalable solution.

cs.CV

版本更新 · v2arXiv:2605.14171

CSI-JEPA: Towards Foundation Representations for Ubiquitous Wi-Fi Sensing with Minimal Supervision

Xuanhao Luo, Zhizhen Li, Yuchen Liu

原文摘要

Channel state information (CSI) provides a widely available sensing modality for human and environment perception, but existing CSI sensing models usually rely on task-specific supervised training and require substantial labeled data for each task, device, user, or environment. This limits their scalability in practical deployments where unlabeled CSI is abundant but labeled data is costly to collect. In this paper, we present CSI-JEPA, a self-supervised predictive representation learning framework for label-efficient and transferable Wi-Fi sensing. CSI-JEPA learns reusable temporal-spectral representations from unlabeled CSI samples by predicting latent features of masked channel regions from visible context. To better match the physical structure of CSI, it jointly tokenizes CSI along the time and subcarrier dimensions and introduces a channel variation-aware masking strategy that prioritizes regions with stronger local channel dynamics. After pretraining, the encoder is frozen and adapted to downstream sensing tasks using lightweight task-specific adapters. We evaluate CSI-JEPA on seven real-world Wi-Fi sensing tasks spanning diverse objectives and deployment settings. The results show that CSI-JEPA improves downstream sensing performance over competitive baselines, achieving up to 10.64 percentage points (pp) mean accuracy gain over the supervised Transformer and matched-budget label savings of up to 95.0%. Across unseen tasks, environments, hardware platforms, and CSI organizations, frozen CSI-JEPA improves downstream adaptation by up to 22.92 pp over non-pretrained baselines, demonstrating strong representation transferability.

cs.LG · cs.NI

版本更新 · v2arXiv:2611.03261

Exact Agreement Expansion up to the Top via Non-polarity

Tali Kaufman, Huy Tuan Pham

原文摘要

Agreement expansion asks whether local assignments to faces of a poset that usually agree on a sampled common face must agree with a single global function on most faces. Agreement expansion is exact (approximate) if local assignments need to exactly (approximately) agree with the global function. We suggest a new proof technique for agreement expansion that enables agreement expansion in settings which were previously beyond reach, such as: exact agreement expansion for Grassmannian posets (before only simplicial complexes support exact agreement expansion); agreement expansion based on assignments to top faces of the a complex (before agreement expansion is generally proven only for assignments to low dimensional faces); agreement expansion from one-sided spectral expansion (before, two sided spectral assumption was needed in general settings). Specifically, for down-closed simplicial and Grassmannian complexes, we prove an exact agreement theorem between any two layers assuming one-sided spectral expansion in links of singletons. As an application, we show that full bases of every rank-$k$ matroid admit exact agreement with soundness $O(k^2/[h(k-h)])$ when tested through $h$-element independent sets. We also obtain exact agreement on Grassmannian posets under one-sided spectral expansion, with stronger soundness under a local distance guarantee. Our key proof ideas center around non-polarity of the distributions of local assignments, probed via a random exposure of elements.

cs.CC · cs.DM · math.CO

版本更新 · v2arXiv:2606.29157

Hardness of Counting Orderings: Taking Gadget Amplification to the Limit

Marcelo Arenas, María Alejandra Schild, Bernardo Subercaseaux

原文摘要

We introduce a new elementary technique for proving $\#P$-hardness results under Turing reductions, and showcase its applicability by settling the counting complexity of several fundamental problems about counting orderings in graphs, posets, and geometric configurations. The technique addresses a natural difficulty in counting reductions: when designing gadgets that attempt to enforce a semantic property, the introduction of the gadgets themselves can affect the function to count in complicated ways. In a nutshell, we introduce an amplification parameter $q$ for the gadgets (e.g., replacing a subset of the vertices of a graph by $q$ copies) and show that the limiting behavior when $q$ tends to infinity, or even to negative values, filters out the unwanted effects of the gadgets, which allows us to recover the desired counts. These limits can be computed symbolically via interpolation, after showing that the probability of a random ordering satisfying the problem's constraints is a rational function in $q$. Concretely, with this methodology, we show $\#P$-completeness for counting the following objects: (a) successive vertex orderings, even on bipartite graphs, (b) $st$-numberings of graphs, (c) shellings, even for bipartite graphs, (d) perfect elimination orderings, even for split graphs, (e) linear extensions of N-free posets of height $3$, (f) linear extensions of posets of height $2$, and (g) peeling sequences of point sets. Result (e) settles a conjecture of Felsner and Manneville (2015), and although result (f) was first proved by Dittmer and Pak (2018), we include an alternative proof, using our technique, that does not rely on the result of Brightwell and Winkler (1991) about hardness of counting for general posets. Result (g) answers a question of Eppstein.

cs.CC · math.CO

版本更新 · v2arXiv:2501.10172

Mean and Variance Estimation Complexity in Arbitrary Distributions via Wasserstein Minimization

Valentio Iverson, Stephen Vavasis

原文摘要

Parameter estimation is a fundamental challenge in machine learning, crucial for tasks such as neural network weight fitting and Bayesian inference. This paper focuses on the complexity of estimating translation $\boldsymbolμ \in \mathbb{R}^l$ and shrinkage $σ\in \mathbb{R}_{++}$ parameters for a distribution of the form $\frac{1}{σ^l} f_0 \left( \frac{\boldsymbol{x} - \boldsymbolμ}σ \right)$, where $f_0$ is a known density in $\mathbb{R}^l$ given $n$ samples. We highlight that while the problem is NP-hard for Maximum Likelihood Estimation (MLE), it is possible to obtain $\varepsilon$-approximations for arbitrary $\varepsilon > 0$ within $\text{poly} \left( \frac{1}{\varepsilon} \right)$ time using the Wasserstein distance.

cs.LG

版本更新 · v2arXiv:2610.34763

The Physical Cost of Context: Quality-Constrained Resource Accounting for Coding Agents

Zhongming Yu, Yitong Ma, Hengjia Yu, Boqin Yuan, Xiangye Lin, Hejia Zhang, Jishen Zhao

原文摘要

Coding agents resend a growing context on every model call, and context policies are usually judged by the prompt tokens they save. Executed work also depends on call count and prefix reuse: a caching runtime prefills only the tokens after the longest cached prefix. A shorter prompt can therefore cost more if it rewrites cached history. Our quality-gated protocol splits executed prefill into logical prompt minus reused prefix and measures KV residency, time, and energy. On 450 localization cases derived from SWE-rebench V2, preloading ten retrieved blocks (Eager) and adding anchored history compaction (Compact) both pass a five-point non-inferiority gate. In a separate 900-cell study, Eager cuts model calls by 35.7% and the GPU-energy proxy by 33.2% against on-demand search (Grep); Compact saves nothing further. Enabling prefix caching leaves Eager's prompt volume similar (+8.3%; its 95% CI spans zero) but cuts its uncached prefill by 67.7%. Deleting history can turn a token saving into extra work. With mini-SWE-agent on 100 SWE-bench Verified instances, keeping only the last five observations (Mask-5) cuts Eager's prompt volume by 33.3% but loses 135.9k reusable tokens per task. Uncached prefill rises by 608.5%, the energy proxy by 125.6%, and the policy misses the quality gate. Policies thus save physical work by shortening the trajectory and can multiply it by editing reusable history.

cs.SE · cs.AI

版本更新 · v2arXiv:2510.02906

FinMultiHopQA: A KG-Grounded Multi-Hop Financial QA Benchmark for Evidence-Driven Reasoning

Abhinav Arun, Reetu Raj Harsh, Bhaskarjit Sarmah, Stefano Pasquali

原文摘要

Multi-hop financial question answering often requires composing evidence distributed across filing sections, fiscal years, and peer companies. Prior financial QA benchmarks address parts of this problem individually, but none jointly support controlled multi-hop taxonomies, document topologies, motif-level reproducibility, and provenance-linked auditability at scale. We introduce FinMultiHopQA, a benchmark of 15,600 qualitative and quantitative questions, grounded in FinReflectKG, a temporally indexed financial knowledge graph derived from SEC 10-K filings with 17.5M+ triples. The pipeline instantiates statistically salient 2-3 hop typed KG motifs across intra-document, inter-year, and cross-company regimes. Beyond closing this gap, the same construction lets us hold each question's underlying evidence constant while varying only its presentation - embedding it in noisy surrounding context or changing its order - giving a controlled instrument to study a secondary question at scale: using six such presentations, we test whether reasoning LLMs' performance depends on evidence structure. Results confirm this: despite their reasoning capacity, judged correctness remains consistently lower under less-structured presentations of the same evidence - most sharply for quantitative cross-company reasoning and 3-hop branch motifs - indicating that reasoning alone does not fully offset how evidence is structured; KG-linked presentation also uses fewer input tokens by design. We release the benchmark, provenance annotations, model outputs, evaluation records, and supporting technical documentation.

q-fin.CP · cs.AI