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

+197新增论文
+151其中 AI 相关
+16已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-21 12:19 · 全站最近入库:2026-11-21 12:19各信源进度 →

统计口径与数据新鲜度

统计 2026-11-21 04:04 — 2026-11-22 04:04(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:2026-11-21 12:19最近发现新增版本:2026-11-21 12:19库内论文最新更新时间:2026-11-20 02:59

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。

29,158 已收录论文24,181 AI 相关8,992 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 8,992 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-21 12:17(北京时间)
版本更新 · v3arXiv:2610.31642

A Fix for Softmax and Cross-Entropy

Jonathan T. Barron, Dan Gnanapragasam, Michael Tschannen

原文摘要

Standard implementations of softmax, log-softmax, and cross-entropy have two numerical bugs. The first bug, Softmax Collapse, rounds the target logit's gradient to zero if that one logit is much larger than the others, resulting in gradients that no longer sum to zero. The second bug, Softmax Cancellation, occurs when student logits are close to teacher logits, and produces non-zero gradients even when the student and teacher logits are equal. We present simple implementations of log-softmax, softmax, and cross-entropy that fix both bugs. Our fix eliminates the previously observed "slingshot" phenomenon, speeds up grokking with weight decay, prevents accuracy collapse without weight decay, stops student networks from drifting away from the teacher during distillation, and shows that after memorization, logit growth is bounded by an equilibrium between weight decay and Adam's $ε$ hyperparameter.

cs.LG

版本更新 · v2arXiv:2609.38866

When Context Changes: Understanding Update Failures in LLMs

Junyu Guo, Yuchen Fang, Shangding Gu, Costas Spanos, James Demmel, Javad Lavaei

原文摘要

As preferences, goals, and facts change, LLM agents must use the current state while earlier versions remain in context. Yet they can answer with an old value of the same variable, a failure that we call stale binding. To study when models use outdated information and why, we introduce Controlled In-Context Memory (CICM), a benchmark for tracking and using updated information in conversations and agent logs. We observe that even frontier reasoning models can fail to recover the current state. We find that in open-source models probes can still recover the updated value when the model answers with an old one, pointing to a failure to select information that remains available. Component tests in Qwen and Pythia identify a mechanism for this selection failure: attention drift, where attention favors old values over the current one when producing an answer. We study a one-layer transformer to mathematically understand how this phenomenon happens: when attention scores are similar, several old values can together receive more attention than the current value. Guided by this explanation, we redirect attention toward the current value without further training. When the current value is requested directly, adjusting this intervention for each input corrects most old-value errors across various model families while preserving nearly all initially correct answers. Reliable context management therefore requires more than remembering updated information: models must use it to guide their answers.

cs.AI

版本更新 · v2arXiv:2610.28629

Mechanistic discovery in grain boundary structure and deformation driven by agentic AI

Yazhuo Liu, Kunqing Ding, Xiaoguo Gong, Zifeng Wang, Yichen Yang, Yifeng Che, Wen Chen, Josh Kacher 等 11 位作者

原文摘要

Agentic artificial intelligence (AI) systems hold immense potential for automating complex workflows in materials discovery and design, but extending these capabilities to domain-intensive mechanistic studies remains challenging. General grain boundaries (GBs) exemplify this challenge, as their structural complexity complicates experimental characterization and atomistic modeling. Here, we present GBClaw, an agentic AI framework for investigating the structure and deformation of general GBs, which are common in real polycrystals yet remain understudied. By integrating large language models into a modular architecture, GBClaw effectively combines systematic simulations with targeted mechanistic analysis through iterative workflow refinement and dynamic replanning. Applied to 14 general asymmetric tilt GBs in nickel, the framework characterized interfacial defect structures and deformation modes across these incommensurate boundaries. It established a general approach to identify previously unobserved, widely spaced secondary GB misfit dislocations (SGBMDs) alongside closely spaced primary GB misfit dislocations (PGBMDs) in simulated atomic configurations and experimental images. Comparisons of GB sliding with and without SGBMDs revealed substantial differences in shear strength, underscoring the critical role of misfit-dislocation hierarchies in sliding resistance. Researcher-guided follow-up calculations further elucidated the structural origins of shear asymmetry and an anomalous shear response. This work demonstrates that combining agentic AI with domain expertise bridges high-throughput computational screening and deep mechanistic discovery at complex material interfaces.

cond-mat.mtrl-sci · cond-mat.mes-hall · cs.AI · cs.MA

版本更新 · v2arXiv:2610.01563

Optimal Universal Coding of Integers

Wei Yan, Yunghsiang S. Han, Leqian Zheng

原文摘要

Universal coding of integers (UCI) provides prefix codes for positive integers such that, for every nonincreasing source distribution $P$, the average codeword length under $P$ stays within $K$ times $\max\{1,H(P)\}$. For a given UCI $\mathcal{C}$, the smallest such constant $K$ is called the minimum expansion factor of $\mathcal{C}$, denoted $C_{\mathcal{C}}^{*}$. Let $C^{*}\triangleq \inf \{C_{\mathcal{C}}^{*}\}$ over all UCIs. A UCI attaining $C^{*}$, or a family of UCIs attaining $C^{*}$ in the limit, is called optimal. It is currently known that $2\le C^{*}\le 2.0386$. In this paper, we construct a family of one-point plus uniform-tail distributions and prove that, for every UCI, a distribution in this family attains the worst-case ratio; thus, the family is least favorable for the UCI problem. We further establish an inequality, called the \emph{UCI inequality}, which plays the same role for UCI as the Kraft inequality does for prefix codes: for any real number $B\ge 1$, it decides whether $B$ lies below or above $C^*$. The UCI inequality gives an equivalent definition of $C^*$. By numerical computation, we determine $C^* = 2.00012475703610197861701176297579886254\cdots$, the first $38$ decimal digits being certified. Once $C^*$ is known, we can theoretically construct the optimal UCI.

cs.IT

版本更新 · v2arXiv:2609.34121

Online Stochastic Allocation with Increasing Returns

Shuo Sun, Yunduan Lin

原文摘要

We study online allocation with increasing returns, where an uncertain number of customers arrive sequentially and can be irrevocably assigned to any product with remaining capacity. Each assignment earns a base reward plus a bonus that is nondecreasing in the number of previous assignments to the same product. Prior work has focused primarily on fixed or diminishing returns. Under increasing returns, the central challenge is that concentrating assignments can unlock large future rewards, while a different product may be preferable if arrivals stop early. We show that classical approaches, including greedy and natural LP-based rounding algorithms, can perform arbitrarily poorly, even with homogeneous bonuses. We evaluate performance through competitive ratios relative to the hindsight offline optimum. For homogeneous bonuses, we establish a tight competitive ratio of $1/2$. For arbitrary heterogeneous bonuses, however, no online algorithm can guarantee a competitive ratio greater than $1/m$, where $m$ is the number of products. By imposing discrete concavity on the bonus functions, we limit the concentration of rewards in later assignments and restore constant guarantees. Under this structure, we develop a re-solving framework that repeatedly selects a planning horizon, constructs a feasible allocation plan from the current state, and executes its assignments in a specific order. For arbitrary arrival distributions, we design distribution-oblivious deterministic and randomized algorithms with competitive ratios of $0.2537$ and $0.3864$, respectively. For known increasing-failure-rate (IFR) arrival distributions, we exploit distributional information to obtain a deterministic $1/2$-competitive algorithm.

cs.DS · math.OC

版本更新 · v2arXiv:2609.05012

Solution-space heterogeneity shapes federated learning dynamics across partial differential equations

Ping Luo, Jiahuan Wang, Ziqing Wen, Kele Xu, Dongsheng Li, Tao Sun

原文摘要

In natural science applications, federated learning (FL) enables collaborative training of models that approximate solutions to partial differential equations (PDEs) for repeated use in distributed settings. However, the input conditions of FL-PDE training samples involve multiple coupled parameters, making data heterogeneity across clients difficult to analyse. We propose solution-space PDE-Dirichlet, which groups existing solution labels into task-specific bins and distributes samples using the classical Dirichlet allocation rule in FL, providing a common partitioning approach across PDE tasks. Theoretical analysis shows an association between differences in solution-bin composition and gradient disagreement, while experiments on three controlled operators and four public simulation datasets quantify the relationships between this heterogeneity, gradient disagreement and prediction error. Based on this connection, we develop Solution-Bin Controlled Federated Optimization, which uses pooled bin-gradient prototypes to correct the component of client gradients associated with bin composition while preserving the sample-weighted training objective. Experiments on low-viscosity Burgers and Darcy flow show that this correction reduces prediction error compared with federated averaging and narrows the gap from bin-balanced IID training.

cs.LG · cs.DC

版本更新 · v4arXiv:2603.22016

ROM: Real-time Overthinking Mitigation via Streaming Detection and Intervention

Xinyan Wang, Xiaogeng Liu, Ming Pei, Chaowei Xiao

原文摘要

Large reasoning models (LRMs) often suffer from overthinking, continuing with redundant verification and repeated attempts even after reaching a correct solution. This wastes computation and can overturn an initially correct answer. We show that the transition from productive to redundant reasoning is decodable from late-layer hidden states. These representations separate the two phases around first-correct-solution boundaries beyond what position and lexical controls explain, and distinguish correct from incorrect solution endings. Building on this finding, we propose ROM (Real-time Overthinking Mitigation), which uses a lightweight streaming detector (0.07--0.25\% of backbone parameters) to monitor a frozen LRM and intervene at well-formed reasoning boundaries without answer extraction or probe decoding. Counterfactual Self-Correction (CSC) augments detector training with wrong$\rightarrow$correct trajectories to preserve useful self-correction. Heads are trained separately for each backbone using shared MATH500 supervision and transfer zero-shot to four unseen benchmarks. Evaluated against ten baselines across five backbones from three families and five benchmarks, ROM$_{\text{CSC}}$ reduces response length by 28--77\% (45\% on average), with significant reductions in all 25 settings and accuracy matching or exceeding vanilla decoding in 22. Measurements on two backbones further show 20--70\% lower end-to-end latency with less than 0.5\% detector overhead per decoding step. Code is available at https://github.com/SaFo-Lab/ROM.

cs.LG · cs.AI · cs.CL

版本更新 · v2arXiv:2610.33947

Learning Functional Specialisation for Bimanual Robotic Control

Jingwen Li, Levin Kuhlmann, Jason Friedman, Gideon Kowadlo

原文摘要

Bimanual robotic control requires simultaneous management of distinct functional demands under motor noise, energetic costs, and communication delays. We investigate how learning responsibility should be organised across modular controllers and the two arms in a simulated musculoskeletal system with two communicating recurrent controllers. Using a 2 x 2 design, we compare differential versus shared controller objectives and fixed versus balanced hand-task assignments. Task-aligned functional specialisation was strongest when differential objectives were combined with a fixed assignment aligned with contralateral control routing. This condition achieved the highest canonical-task performance and control efficiency, reaching 81.33% bimanual success. Its advantage over shared objectives increased with motor noise but narrowed under stronger energy penalties. Stronger specialisation also reduced hand-role flexibility: reversing the hand roles decreased bimanual success by 80.46 percentage points under Differential-Fixed training, compared with 1.51 percentage points under Shared-Balanced training. These results show that controller objectives and hand-task experience jointly determine functional organisation. Stronger role-specific specialisation favours performance and efficiency when roles remain stable, whereas more distributed training favours interchangeability when roles change.

cs.RO

版本更新 · v3arXiv:2609.36485

From Reconnaissance to Response: Quantitative Risk Parameterization and Game Theoretic Containment in Modern Enterprise Attack

Shadeeb Hossain, Elias Calboreanu

原文摘要

Modern Security Operations Centers struggle with delayed manual incident response, enabling adversaries to advance through the Cyber Kill Chain during early stage reconnaissance. While classical game theoretic defense models optimize strategic resource allocation, they rely on static utility matrices that fail to adapt to dynamic telemetry. This paper presents an integrated, metrics driven decision engine that bridges quantitative risk parameterization and continuous automated response time. Common Vulnerability Scoring Systems exploitability parameters are mapped to attacker success probabilities and evaluate defender log distributions via Factor Analysis of Information Risk Monte Carlo simulations. Real time SIEM logs streams are modeled as Poisson process arrival rates, dynamically updating defender posterior threat belief through sequential Bayesian filtering. A closed form threshold is derived by framing the interaction as a dynamic Bayesian Stackelberg game, where the expected unmitigated risk exceeds proactive containment cost. Parameterized against empirical data from the 2023 MGM Resorts and Caesars Entertainment cyber incident, simulation results demonstrate that the engine suppresses transient background noise while triggering automated SOAR network isolation within seconds of adversarial probing. Multi parameter sensitivity analysis confirms that the decision boundary dynamically adjusts to live perimeter vulnerability, offering a control theoretic foundation for sub minute automated threat containment.

cs.GT · cs.CR · eess.SY

版本更新 · v3arXiv:2510.02219

Contrastive Retrieval Heads for Improved Attention-Based Reranking

Linh Tran, Yulong Li, Radu Florian, Stacy Patterson, Ana Milanova, Wei Sun

原文摘要

The strong zero-shot and long-context capabilities of recent Large Language Models (LLMs) have paved the way for highly effective re-ranking systems. Attention-based re-rankers leverage attention weights from transformer heads to produce relevance scores, but not all heads are created equally: many contribute noise and redundancy, thus limiting performance. To address this, we introduce CoRe heads, a small set of retrieval heads identified via a contrastive scoring metric that explicitly rewards high attention heads that correlate with relevant documents, while downplaying nodes with higher attention that correlate with irrelevant documents. This relative ranking criterion isolates the most discriminative heads for re-ranking and yields a state-of-the-art list-wise re-ranker. Extensive experiments with three LLMs show that aggregated signals from CoRe heads, constituting less than 1% of all heads, substantially improve re-ranking accuracy over strong baselines. We further find that CoRe heads are concentrated in middle layers, and pruning the computation of final 50% of model layers preserves accuracy while significantly reducing inference time and memory usage.

cs.IR

版本更新 · v2arXiv:2608.07520

KumbhDoot: A Scale-Ready, LLM-Bounded Architecture for Mass-Gathering Public-Service Assistants

Saurabh Sakalkar, Abhishek Singh, Ramesh Raskar

原文摘要

Mass religious gatherings such as the Kumbh Mela concentrate tens of millions of people into a single region over a few weeks, producing intense, repetitive, multilingual, and safety-critical demand for information. The default response, a conversational assistant that routes every query to a large language model (LLM), is poorly matched to this setting: it is costly at scale, slow on emergency paths, prone to hallucination on facts that can cause physical harm, and unusable when connectivity fails. We describe KumbhDoot, an agentic pilgrim assistant for the Nashik Simhastha Kumbh Mela built on a different principle. It operates on a foundational design principle that prioritizes semantic similarity over starting with an LLM. Generative models are invoked only in instances where similarity-based retrieval is insufficient to produce a correct answer. The system utilizes a "semantic cache": an embedding-indexed store as a single retrieval primitive, which handles intent routing, answer caching, offline lookups, and multi-agent retrieval. A custom three-tier agent architecture operates directly on this store, ensuring decision paths remain inspectable and avoiding the use of generic multi-agent frameworks that would trigger implicit per-step LLM calls. We present the architecture, an analytical cost model for its per-query economics, and an honest account of where similarity is sufficient and where generative reasoning remains necessary. We argue that for bounded, high-stakes, low-connectivity public-service domains, a similarity-first and LLM-bounded design is not merely cheaper but architecturally more appropriate than an LLM-default one.

cs.CY · cs.AI

版本更新 · v2arXiv:2604.16477

A Constructive Proof of Rice's Theorem and the Halting Problem via Hilbert's Tenth Problem

Jonathan Brossard

原文摘要

Rice's theorem states that no non-trivial semantic property of programs is decidable. Classical proofs proceed by reduction from the halting problem, invoking the law of excluded middle (LEM) twice: once through diagonalization, and once through a case split on whether the always-diverging program bot satisfies the property in question. We present a proof that is constructive relative to the undecidability of Hilbert's Tenth Problem (MRDP): valid in intuitionistic logic, requiring neither diagonalization nor self-reference, and adding no classical reasoning beyond the MRDP assumption itself. The key idea is a two-witness construction. Given a non-trivial property P, we attach to each Diophantine polynomial D a pair of programs S^0_D, S^1_D that behave like the negative and positive witnesses for P when D is solvable, and both diverge identically when it is not. A hypothetical decider for P would therefore decide Diophantine solvability via the difference delta_D = DecideP(S^1_D) - DecideP(S^0_D) -- contradicting the MRDP theorem. The argument is structured as two separate implications, never asserting a disjunction about solvability, and never examining P(bot). The undecidability of the halting problem follows as an immediate corollary: a single application of Rice's theorem to the Terminates property. A formalization in the Rocq proof assistant confirms both results within a step-indexed model of computation, with the undecidability of Hilbert's Tenth Problem as the sole external axiom. Both Rice_Theorem and Halting_Problem are closed under the global context.

cs.LO · cs.CR

版本更新 · v4arXiv:2605.19811

LionMuon: Alternating Spectral and Sign Descent for Efficient Training

Arman Bolatov, Artem Riabinin, Nikita Kornilov, Andrey Veprikov, Samuel Horváth, Martin Takáč, Aleksandr Beznosikov

原文摘要

Muon's spectral step gives a stronger direction than a sign step, but it is expensive. Every step runs Newton-Schulz iterations on the full matrix and, in distributed training, an extra all-reduce. Sign steps, as in Lion and Signum, are cheap and stay local to each device. We propose LionMuon, which takes one Muon step every P iterations and Lion steps in between, with a single dual-EMA momentum buffer shared by both. Muon's compute and communication are paid once per P steps, and the optimizer state is one buffer per matrix, half of AdamW's. A single-EMA variant, SignMuon, isolates the effect of alternation. We prove complexity bounds under heavy-tailed noise in which the period interpolates between the smoothness and noise constants of Muon and Lion, and which state, through measurable ratios of these constants, when LionMuon's complexity is below Muon's. On 124M and 355M models trained on FineWeb, LionMuon with P=2 and P=5 reaches a lower loss than Muon, AdamW, Lion and Signum at the same number of tokens. Under 4-GPU data-parallel training LionMuon reaches Muon's final loss with up to 30% less wall-clock on PCIe, and it beats the communication-efficient Muon variants Dion and MuonBP on loss at no more exposed communication, while keeping the exact gradient. Code: https://github.com/brain-lab-research/lion-muon

cs.LG

版本更新 · v2arXiv:2605.01097

Interpretable Difficulty-Aware Knowledge Tracing in Tutor-Student Dialogues

Shuyan Huang, Alexander Scarlatos, Jaewook Lee, Andrew Lan

原文摘要

Recent advances in large language models (LLMs) have led to the development of AI-powered tutoring systems that provide interactive support via dialogue. To enable these tutoring systems to provide personalized support, it is essential to assess student performance at each turn, motivating knowledge tracing (KT) in dialogue settings. However, existing dialogue-based KT approaches often ignore question difficulty modeling and rely on opaque latent representations from LLMs, hindering accurate and interpretable prediction. In this work, we propose an interpretable difficulty-aware conversational KT framework built upon LLMs, which explicitly models students' abilities and the difficulty of tutor-posed tasks at each turn. The framework incorporates the original textual question and the next tutor-posed task to estimate the student's knowledge state and the difficulty of the upcoming turn. Furthermore, it integrates Item Response Theory to map LLM's outputs into student ability and question difficulty parameters, enabling interpretable prediction of student performance grounded in cognitive theories of learning. We evaluate the framework on two tutor-student dialogue datasets. Both quantitative and qualitative results show that our framework outperforms existing KT baselines, meanwhile generating interpretable outputs consistent with cognitive theory.

cs.CL · cs.AI

版本更新 · v3arXiv:2606.14867

Evaluating the Robustness of Proof Autoformalization in Lean 4

Zhengtao Gui, Sheng Yang, Zhouxing Shi

原文摘要

Proof autoformalization aims to translate a mathematical informal proof in natural language into a formal proof in a formal language such as Lean 4. Existing LLM-based models for proof autoformalization are typically evaluated only on well-formed informal proofs from curated datasets. We argue that a robust proof autoformalizer must remain faithful even for informal proofs that diverge from these idealized ones, and we present the first study on the robustness of proof autoformalization models. We formulate two categories of perturbations: a global perturbation paraphrases the informal theorem-proof pair in a different style, under which the formalization should remain consistent; a local perturbation alters a value, symbol, or proof step, possibly in a counterfactual way, and a robust formalization should faithfully reflect the perturbation rather than reverting to the original or inferring a different one on its own. We build a benchmark with both perturbations on miniF2F and MATH-500, and automatically measure how stable a proof autoformalization's correctness is under global perturbations and how faithfully its output reflects local perturbations. We evaluate seven recent models: their correctness varies under global perturbations, and they mostly fail to remain faithful under local perturbations.

cs.CL · cs.AI · cs.LG

版本更新 · v2arXiv:2604.07424

An Analysis of Artificial Intelligence Adoption in NIH-Funded Research

Navapat Nananukul, Mayank Kejriwal

原文摘要

Understanding the landscape of artificial intelligence (AI) and machine learning (ML) adoption across the National Institutes of Health (NIH) portfolio is critical for research funding strategy, institutional planning, and health policy. The advent of large language models (LLMs) has fundamentally transformed research landscape analysis, enabling researchers to perform large-scale semantic extraction from thousands of unstructured research documents. In this paper, we illustrate a human-in-the-loop research methodology for LLMs to automatically classify and summarize research descriptions at scale. Using our methodology, we present a comprehensive analysis of 58,746 NIH-funded biomedical research projects from 2025. We show that: (1) AI constitutes 15.9% of the NIH portfolio with a 13.4% funding premium, concentrated in discovery, prediction, and data integration across disease domains; (2) a critical research-to-deployment gap exists, with 79% of AI projects remaining in research/development stages while only 14.7% engage in clinical deployment or implementation; and (3) health disparities research is severely underrepresented at just 5.7% of AI-funded work despite its importance to NIH's equity mission. These findings establish a framework for evidence-based policy interventions to align the NIH AI portfolio with health equity goals and strategic research priorities.

cs.AI · cs.CY · cs.MA

版本更新 · v2arXiv:2607.01033

The Model Organism Lottery: Model Organism Interpretability Strongly Depends on Training Methodology

Andrzej Szablewski, Gabriel Konar-Steenberg, Raffaello Fornasiere, Nikita Menon, Stefan Heimersheim

原文摘要

Model organisms (MOs) - language models trained to exhibit undesired or unnatural behaviours - are frequently used as testbeds for evaluating white-box interpretability techniques. Current MOs are typically constructed via post-hoc supervised fine-tuning (SFT). Prior research has shown that interpretability methods can easily identify hidden behaviours in these MOs. However, recent work suggests that such post-hoc training methods may make interpretability unrealistically easy. We investigate this claim by constructing a suite of 61 $\verb|OLMo2-1B|$- and $\verb|gemma-3-1b-it|$-based MOs trained with seven different techniques, including standard post-hoc SFT, post-hoc DPO, and more realistic integration of MO data into the OLMo post-training DPO phase. We also use these to train 96 further variants via knowledge distillation. We use these MO variants to benchmark activation oracles, activation steering, the logit lens, and the Jacobian lens. Our findings show that (i) MO interpretability depends strongly on training objective, target behaviour, model architecture, and training data generation pipeline; (ii) substantial variance remains even after controlling for differences in the strength of target behaviour expression; and (iii) both our more realistic $\textit{integrated training}$ and knowledge distillation often yield less interpretable MOs than standard post-hoc methods. Our results cast substantial doubt on the validity of current MOs as interpretability proxies and suggest actionable steps towards more robust benchmarks.

cs.LG

版本更新 · v2arXiv:2603.22691

Rank-Aware CPU Reservation and In-Place Adjustment for MPI CFD on Kubernetes

Tianfang Xie

原文摘要

Non-uniform meshes in computational fluid dynamics (CFD) can leave some Message Passing Interface (MPI) ranks waiting while others determine execution time. On Kubernetes, reserving a full CPU for every rank can then restrict the admission of further simulations. We investigate rank-aware CPU reservation and in-place request adjustment using OpenFOAM on three Kubernetes testbeds. The framework lowers reservations without hard CPU caps and uses online MPI profiles to adjust each rank's CPU request while the solver runs. We reduced the per-job CPU reservation from 16.0 to 5.9 vCPU, with single-job wall-clock time increasing by at most 6%. With these reservations on twelve nodes, six co-located simulations reached 4.09 times the throughput of sequential execution. Matched controls showed no consistent benefit from proportional request shaping. Beyond the nominal CPU capacity, throughput collapsed with default busy-waiting; yielding waits softened the collapse. In one six-simulation run, the online controller performed 120 in-place resizes of running solvers with zero pod restarts; the pipeline reached 2.5 times the throughput of sequential execution. In a separate comparison on 16 physical cores, reallocating ranks from the far field to the near-wall zone reduced fixed-iteration wall-clock time by factors of 2.0 to 4.2 relative to the original concentric allocation, depending on mesh and iteration horizon. The reallocation also outperformed balanced Scotch partitions. The implementation requires no changes to OpenFOAM, the MPI library, or the Kubernetes scheduler.

cs.DC

版本更新 · v2arXiv:2606.04262

Can I Take Another Dose? DoseBench: Evaluating LLM Decision-Making in OTC Dosing QA

Maroof Kousar, Jinsung Han, Mohit Chandra, Munmun De Choudhury, Yibo Hu

原文摘要

Large language models (LLMs) are increasingly used to answer everyday health questions online, including whether someone can take another dose of an over-the-counter (OTC) medication. Yet this common safety-relevant setting remains underexplored in medical QA. Answering these questions requires checking previous doses, timing, and product-specific dosing limits, while recognizing when important information is missing. We introduce DoseBench, a focused benchmark of 230 consumer-style acetaminophen and ibuprofen dosing questions with human-annotated YES, NO, and AMBIGUOUS labels, and evaluate eight LLMs with five responses per question. Overall accuracy ranges from 44.8% to 62.2%, while models incorrectly approve 14.0% to 60.5% of prohibited doses. Models with similar accuracy can therefore make very different dosing decisions. Additional analyses show that these failures are not easily fixed by giving models more guidance. Asking models to check dose, timing, and missing information makes them more cautious, while providing drug facts improves accuracy but can make them overlook missing information. These findings highlight a broader challenge in health QA: applying rules correctly and recognizing when information is missing.

cs.CL · cs.AI

版本更新 · v2arXiv:2604.03529

Incentives Shape How Humans Co-Create with Generative AI

Nathanael Jo, Manish Raghavan

原文摘要

Generative AI is quickly becoming an integral part of people's everyday workflows. Early evidence has shown that while generative AI can increase individual-level productivity, it does so at the cost of collective diversity, potentially narrowing the set of ideas and perspectives produced. Our research stands in contrast to this concern: through two pre-registered randomized control trials, we show that incentives can moderate AI's homogenizing force in two creative writing tasks where participants can use AI interactively. Participants rewarded for originality relative to peers produce collectively more diverse writing than those rewarded for quality alone. These participants continued to engage with AI, but ultimately incorporated less AI-generated text verbatim. Our results reveal that the effects of generative AI depend not only on the technology itself, but also the behavioral strategies and incentive structures surrounding its use. We release a rich, anonymized dataset capturing fine-grained trajectories of human-AI co-creation across both studies.

cs.HC · cs.AI · cs.CY

版本更新 · v3arXiv:2608.10319

Do Personalized Skills Help Coding Agents? An Empirical Study of Developer Interaction Traces

Shuyan Huang, Kai Du, Andrew Lan

原文摘要

As human developers increasingly collaborate with large language model (LLM)-powered coding agents, adapting to individual developer preferences, potentially via agent skills, may offer way to reduce repeated guidance and improve collaboration efficiency. However, existing approaches primarily focus on building skills for each task rather for each human developer. In this work, we propose a framework that combines rule-based bootstrapping with LLM-based refinement to distill developer-specific skills from their interactions with coding agents. We further develop an interactive replay framework with an LLM-based developer simulator for evaluation. We conduct experiments on a collection of replayable real-world developer-agent interaction sessions. We find that overall, task-specific skills are the most effective; However, personalized skills become more effective when developer preferences are manifested with enough frequency, especially when developers work on tasks that are highly related. These findings provide empirical insights suggest that i) broadly transferable procedural knowledge can be more robust than developer-specific preference signals and ii) effective personalization requires mining from rich user preference signals.

cs.SE · cs.AI

版本更新 · v2arXiv:2602.00414

Toward Autonomous Laboratory Safety Monitoring with Vision Language Models: Learning to See Hazards Through Scene Structure

Trishna Chakraborty, Udita Ghosh, Aldair Ernesto Gongora, Ruben Glatt, Yue Dong, Jiachen Li, Amit K. Roy-Chowdhury, Chengyu Song

原文摘要

Laboratories are prone to severe injuries from unsafe actions, yet continuous manual safety monitoring, beyond mandatory pre-lab safety training, is infeasible. Vision-language models (VLMs) offer promise for autonomous laboratory safety monitoring, but their effectiveness in realistic settings is unclear due to the lack of visual evaluation data. Safety incidents are mostly documented as text, and existing studies are largely limited to textual understanding. Since hazardous scenes are unsafe to stage or capture, synthetic visual data becomes the only practical way to study this problem. This motivates us to introduce a structured data generation pipeline that converts textual laboratory scenarios into aligned quadruples using LLM as scene graph architect and image generation model as renderer, yielding \dataset: a synthetic dataset of 1,207 $\langle$image, scene graph, ground truth, hazard relevance mask$\rangle$ quadruples across 362 unique scenarios. Experiments on \dataset and seven open- and closed-source models show that VLMs perform effectively given a textual scene graph but degrade substantially in visual-only settings (i.e., where no scene graph is involved), indicating difficulty in extracting structured object relationships directly from pixels. We thus propose \saga, a post-training framework that leverages reinforcement learning to first structure the visual scene as a scene graph, and then predict hazards while aligning attention with relevant image regions. Evaluations with Qwen 3.5 and Gemma 4 show that our method nearly doubles the $F_1$ score over the baseline on our synthetic dataset and achieves an improvement of $\sim 3\times$ on a mini real-world dataset, while preserving general visual understanding in visual question answering.

cs.CV · cs.LG

版本更新 · v4arXiv:2505.00311

A Practical GPU-Enhanced Matrix-Free Primal-Dual Method for Large-Scale Conic Programs

Zhenwei Lin, Zikai Xiong, Dongdong Ge, Yinyu Ye

原文摘要

In this paper, we introduce a practical GPU-enhanced matrix-free first-order method for solving large-scale conic programming problems, which we refer to as PDCS, standing for the Primal-Dual Conic Programming Solver. Problems that it solves include linear programs, second-order cone programs, convex quadratic programs, and exponential cone programs. The method avoids matrix factorizations and uses sparse matrix-vector multiplication as its core operation, which is memory-efficient and well suited for GPU acceleration. It builds on a variant of the primal-dual hybrid gradient method (PDHG) called reflected restarted Halpern PDHG with averaging, which we prove converges on general conic programs, and further incorporates several practical enhancements. Its GPU implementation, cuPDCS, uses customized computational schemes that assign cones of different types and sizes to different levels of the GPU architecture. Numerical experiments demonstrate that cuPDCS is generally more efficient than state-of-the-art commercial solvers and other first-order methods on large-scale conic program applications, including Fisher market equilibrium problems, Lasso regression, and multi-period portfolio optimization. cuPDCS also exhibits better scalability, efficiency, and robustness compared to other first-order methods on the conic program benchmark dataset CBLIB. These advantages are more pronounced in large-scale settings.

math.OC · cs.MS

版本更新 · v2arXiv:2610.25752

Which Answers Agree? Evaluating Self-Consistency from Incomplete Logs

Zhengtao Gui, Zexi Chen, Weining Shen

原文摘要

Self-consistency answers a question by generating multiple responses and selecting the most frequent answer. Its evaluation requires both correctness labels and information about which responses agree; correctness labels alone generally determine neither voting accuracy nor the benefit of additional votes. For tasks in which all correct responses map to the same normalized answer, we derive sharp accuracy bounds using classical stochastic majorization, refined by abstention counts and answer-category limits. We also derive sharp bounds on accuracy changes across voting budgets that share a response pool. When answer-category counts are available, we introduce Plurality-U, a count-based dynamic program that exactly averages voting accuracy over all fixed-size subsets, eliminating finite-pool Monte Carlo error. On 1,403,520 stored outputs from recent reasoning models in Scorio Lite and Scorio Trace, recording abstention counts and answer options cuts the width of SuperGPQA accuracy intervals by more than half. For moving from 9 to 17 votes, our bounds reduce interval width by a median of 77.7% on SuperGPQA and 72.3% on 14 GPQA and Adaptive-Consistency pools, compared with a baseline that combines separate accuracy bounds. Exact computation has lower paired median runtime than 2,000-subset Monte Carlo in 14 of 15 tested pool-budget settings. These results clarify what incomplete logs can establish and provide an exact evaluation procedure when answer counts are retained.

cs.LG

版本更新 · v3arXiv:2510.00361

Attribution Gradients: Incrementally Unfolding Citations for Critical Examination of Attributed AI Answers

Hita Kambhamettu, Alyssa Hwang, Philippe Laban, Andrew Head

原文摘要

AI answer engines are a relatively new kind of information search tool: rather than returning a ranked list of documents, they generate an answer to a search question with inline citations to sources. But reading the cited sources is costly, and citation links themselves offer little guidance about what evidence they contain. We present attribution gradients, a technique to boost the informativeness of citations by consolidating scent and information prey in place. Its first feature is bringing evidence counts, supporting/contradictory excerpts, links to sources, and contextual explanations into one place. Its second feature is the ability to unravel second-degree citations in place. In a lab study, we demonstrate usage of the full gradient in a critical reading task and its support for deep engagement that increased the depth of what readers took away from the sources versus a standard citation and document QA design.

cs.HC · cs.AI

版本更新 · v7arXiv:2509.25783

Sharpness of Minima in Deep Matrix Factorization

Anil Kamber, Rahul Parhi

原文摘要

Understanding the loss-landscape geometry near a minimum is key to explaining the implicit bias of gradient-based methods in non-convex optimization problems such as deep neural network training and deep matrix factorization. A central quantity for characterizing this geometry is the maximum eigenvalue of the Hessian of the loss. Currently, its precise role has been obfuscated because no exact expressions for this sharpness measure were known in general settings. In this paper, we present the first exact expression for the maximum eigenvalue of the Hessian of the squared-error loss at any minimizer in deep matrix factorization problems, resolving an open question posed by Mulayoff and Michaeli [2020]. This expression reveals a fundamental property of the loss landscape in deep matrix factorization: It is sufficient, but not necessary, for flatness that, across all layers, the product of the spectral norms of the left and right intermediate factors equals a constant determined by the depth and the spectral norm of the target matrix. Most notably, in both depth-2 matrix factorization and deep overparameterized scalar factorization, we show that this condition is both necessary and sufficient for flatness, which implies that flat minima are spectral-norm balanced and not necessarily balanced in depth-2 matrix factorization. To complement our theory, we provide empirical verification, relying on our exact expression, of an escape phenomenon during gradient-based training near a minimizer of a deep matrix factorization problem.

stat.ML · cs.LG · math.OC

版本更新 · v2arXiv:2610.22686

Multimodal Filesystem Memory For Long-Horizon Embodied Agents

Chloe Su, Jiahan Li, Zhenting Qi, Chaoran Cheng, Haoyu Chen, Sejoon Jun, Weirui Ye, Yilun Du

原文摘要

Agents operating over long horizons need an effective memory that allows them to reuse their experience throughout long interactions. If the entire history of an interaction is retained, it becomes difficult to find the relevant experience, while compressing it into summaries may result in the loss of visual and spatial details that are required later on. We present \textbf{MFM}, a multimodal filesystem memory architecture that keeps complementary types of agent memory. \textbf{Episodic memory} stores the observations, actions, and outcomes, and \textbf{spatial memory} associates experiences with locations. \textbf{Semantic and procedural memory} build up general knowledge and reusable methods of acting, and \textbf{working memory} holds the current plans. MFM makes use of searchable textual records to connect these different representations to the underlying multimodal episodes. The agents are able to find previous experiences by means of contextual cues and to examine the original observations in order to recover details that are missing from their textual descriptions. Through consolidation, the agents convert experience into higher-level knowledge and procedures while still keeping references to the original episodes. In the case of long-horizon Minecraft and computer-use tasks, MFM leads to improved performance with three base models. With GPT-6-Astra, MFM completes \emph{Mine Diamond from Scratch} and reaches 16 out of 24 milestones on the \emph{Ender Dragon} task, as against 12.7 for the best baseline. On CUABench-KiCad it attains a mean reward of 0.440, as compared to 0.400 for the best baseline and 0.267 in the case of no persistent memory.

cs.CL

版本更新 · v3arXiv:2604.19673

InHabit: Scaling 3D Human-Scene Interaction Data With Image Foundation Models

Nikita Kister, Pradyumna YM, István Sárándi, Jiayi Wang, Anna Khoreva, Gerard Pons-Moll

原文摘要

Embodied AI and human-centric 3D scene understanding require data of people meaningfully interacting with diverse environments, yet such data remains scarce. Real-world capture has steadily pushed toward more scenes, but is bound to studios and at most around a hundred scenes, while synthetic datasets scale easily but place people without regard to what the scene affords. We argue that useful human-scene interaction data requires diversity in both scenes and interactions, and we show this with our proposed data engine InHabit. InHabit is a fully automatic engine that renders an existing 3D scene, uses 2D foundation models to propose and depict context-appropriate interactions, and lifts them into SMPL-X bodies, where the known scene geometry makes lifting well-constrained. Applied to Habitat-Matterport3D, InHabit produces InHabitants, which pairs photorealistic images with SMPL-X bodies and complete scene geometry across ~6.6k rooms, about 60x more scenes than the largest captured dataset. We illustrate the quality and value of such large-scale human-scene interaction data through a variety of experiments. In a perceptual study, our interactions are preferred over two prior placement methods in 78% of cases. Across three tasks - 1) contact estimation, 2) human-scene reconstruction, and 3) generating humans in scenes - training with InHabitants outperforms training on synthetic data with exact ground truth (BEDLAM), motion-capture data (TRUMANS), and real RGB-D captures (PROX); e.g., GRAFT trained on InHabitants reaches a contact F1 of 0.594 on PROX, versus 0.492 when trained on PROX itself. Project page: https://virtualhumans.mpi-inf.mpg.de/inhabit/

cs.CV

版本更新 · v5arXiv:2609.01306

MeshSuite: A Unified Benchmark for Triangle- and Mesh-Based Neural Rendering

Kaixuan Zhang, Minxian Li, Mingwu Ren, Xiatian Zhu

原文摘要

Triangle- and mesh-based neural rendering seeks to bridge neural representations and graphics engines through triangle primitives compatible with standard rasterization. Yet existing methods were developed and evaluated under inconsistent settings, with few direct comparisons and little evaluation in graphics engines. We introduce MeshSuite, the first benchmark for systematic evaluation of engine deployment and Non-Engine Render in triangle- and mesh-based neural rendering. MeshSuite evaluates two deployment settings: Standard Deployment, using an opaque mesh pipeline with vertex colors and hardware Z-buffering; and Dedicated Deployment, using method-specific engine implementations to preserve appearance and compositing features. For mesh splatting, we also audit exported surfaces for topological and geometric integrity as downstream graphics assets. Our results show that engine deployment reduces image quality in both settings; among evaluated methods, MeshSplatting degrades least under Standard Deployment. Dedicated Deployment preserves most fidelity at roughly 5-33x the runtime of Standard Deployment across two deployment datasets. For mesh splatting, explicit connectivity and shared indexing remain limited: shared vertex indexing alone guarantees neither manifoldness nor global connectivity. These findings show that rasterizability is only part of graphics readiness and underscore the need to assess engine deployment. Project page: https://x-up-lab.github.io/research/meshsuite.

cs.GR · cs.CV

版本更新 · v4arXiv:2512.12078

The Procedural Semantics Gap in Structured CTI: A Measurement-Driven STIX Analysis for APT Emulation

Ágney Lopes Roth Ferraz, Sidnei Barbieri, Wagner Comin Sonaglio, Lourenço Alves Pereira Júnior

原文摘要

Cyber threat intelligence (CTI) encoded in STIX and structured according to the MITRE ATT&CK framework has become a global reference for describing adversary behavior. However, ATT&CK was designed as a descriptive knowledge base rather than a procedural model. We therefore ask whether its structured artifacts contain sufficient behavioral detail to support multi-stage adversary emulation. Through systematic measurements of the ATT&CK Enterprise bundle, we show that campaign objects encode just fragmented slices of behavior. Only 35.6% of techniques appear in at least one campaign, and neither clustering nor sequence analysis reveals any reusable behavioral structure under technique overlap or LCS-based analyses. Intrusion sets cover a broader portion of the technique space, yet omit the procedural semantics required to transform behavioral knowledge into executable chains, including ordering, preconditions, and environmental assumptions. These findings reveal a procedural semantic gap in current CTI standards: they describe what adversaries do, but not exactly how that behavior was operationalized. To assess how far this gap can be bridged in practice, we introduce a three-stage methodology that translates behavioral information from structured CTI into executable steps and makes the necessary environmental assumptions explicit. We demonstrate its viability by instantiating the resulting steps as operations in the MITRE Caldera framework. Case studies of ShadowRay and Soft Cell show that structured CTI can enable the emulation of multi-stage APT campaigns, but only when analyst-supplied parameters and assumptions are explicitly recorded. This, in turn, exposes the precise points at which current standards fail to support automation. Our results clarify the boundary between descriptive and machine-actionable CTI for adversary emulation.

cs.CR