跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+0新增论文
+0其中 AI 相关
+0已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-22 08:44 · 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 12:45 — 2026-11-22 12:45(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:2026-11-22 08:44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

浏览论文库

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 23,345 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-22 08:42(北京时间)
新论文 · v1arXiv:2611.25862

HALTT: A HAte Lexicon Through Time

Dario Franzese, Farah Benamara, Véronique Moriceau, Viviana Patti

原文摘要

Automated hate speech detection degrades over time as everyday vocabulary is weaponized into covert slurs while retaining original benign meanings (polysemy retention). Existing lexical semantic change detection (LSCD) frameworks struggle with this phenomenon due to a lack of standardized, sense-delineated diachronic resources. We propose a multi-stage framework to construct and validate a diachronic lexicon of semantically shifted hate speech terms. Filtering > 20,000 candidates, our pipeline combines historical (1770-1964) and modern (2001-2024) corpus frequency grounding, masked definition-level scoring to mitigate LLM temporal bias, and a dual-persona reasoning prompt paradigm. We release HALTT, a grounded lexicon of 19 polysemous terms with literal and derogatory senses, analyze shift typologies, and establish evaluation protocols for downstream diachronic benchmarking.

cs.CL

版本更新 · v4arXiv:2605.07671

The Endogeneity of Miscalibration: Impossibility and Escape in Scored Reporting

Lauri Lovén, Sasu Tarkoma

原文摘要

An agent's probability report is paid for twice: by a strictly proper scoring rule, which rewards accuracy, and by an approval rule for the decision it triggers. An approval rule that is not affine is known to defeat truthful reporting. We show the conflict is built in: whenever perfect screening is attainable, the welfare-maximizing approval rule is never affine. The distortion is predictable: there is a reserve report at which pretending to be the marginal type costs exactly what approval is worth. Approving every report at or above the reserve sorts agents as if their beliefs were observed, under every strictly proper score and for any belief distribution. So does a continuous rule with one corner; under strict feasibility no continuously differentiable rule does. The binding constraint is steepness, not smoothness: a slope-capped rule sorts perfectly if and only if the cap reaches the critical slope, the steepest chord of the pretending cost up to the reserve; below it the welfare loss is cubic in the shortfall. For the Brier, log and power scores, and whenever the pretending cost is convex up to the reserve, the critical slope has a closed form. The instances are AI-agent oversight and marketplace operation.

cs.GT · cs.AI · cs.MA · econ.TH · math.OC

新论文 · v1arXiv:2611.25861

Decoupled Knowledge Retrieval and Compact Memory Enable Reliable Lifelong Model Editing

Yue Che, Jingze Wang, Chen Guoyi, Yujie Wu, Xuming Ran, Mingkun Xu

原文摘要

Lifelong model editing requires continuously incorporating new facts into deployed large language models (LLMs) without expensive retraining or compromising existing capabilities. Memory-based editors offer an appealing solution by keeping the base parameters untouched, yet their practical use is hindered by linearly growing storage requirements and deteriorating retrieval performance as edits accumulate. To tackle these problems, we introduce BiSpace, a lifelong model editing framework that decouples memory retrieval from knowledge injection across two distinct representation spaces. BiSpace routes queries and identifies relevant edits in an external semantic space, while injecting the corresponding knowledge into the internal representation space of the LLM through sparse and storage-efficient updates. Extensive experiments show that BiSpace shrinks each entry's storage to only about 10.2% of the original and shows strong resilience to long edit sequences of up to 16K edits. On a more challenging benchmark, BiSpace further improves Overall Performance by more than 23% over existing methods. These results advance memory-based editing toward reliable long-term model maintenance at scale.

cs.AI

新论文 · v1arXiv:2611.25860

Neckpinches in a Positively Curved Market: Ricci Flow with Surgery on the Nikkei-225

Siddharta Laloux, Bhargavi Srinivasan

原文摘要

We study the intrinsic geometry of the Nikkei-225 through the discrete Ollivier-Ricci curvature (ORC) of its empirical correlation graph, using the graph Ricci flow and surgery methodology of arXiv:2510.15942. The 226-node correlation graph, built from daily returns between January 2020 and June 2026, is ORC-spherical: every edge has strictly positive curvature before any flow, close to the equal-weight value $κ\approx0.502$. Under a fixed 20-iteration Ricci flow the curvature distribution polarises into a high-curvature bulk and a low-curvature minority, and cutting the low-curvature edges splits the graph into components of 184 and 42 nodes. The recursive ARCUS procedure (Adaptive Ricci Curvature Unsupervised Surgery) stops the flow at each node once a curvature gap ratio $G$ exceeds $G_{\min}$ and the surgery set is stable. With $G_{\min}=10$ and a minimum cluster size of 15, ARCUS performs 28 surgeries over 13 levels and ends in 29 terminal clusters of 2-14 nodes. The stopping rule fires at every node, with $G$ between 10.0 and 92.4, and the top-level split reproduces the fixed-schedule 184/42 partition exactly. Eight small subgraphs stop with $G<15$, so the deeper levels are conditional on $G_{\min}$. Many terminal clusters are single industries by inspection, among them nine banks, five insurers, five general trading companies, three steelmakers and three shipping lines. Two delisted stocks without price data enter as zero-correlation placeholders; with them excluded, every terminal cluster is more strongly correlated internally than with the rest of the index.

q-fin.ST · cs.SI · physics.soc-ph

新论文 · v1arXiv:2611.25859

Optimal Impure $(r,δ)$-Quantum Locally Recoverable Codes with Explicit Minimum Localities

Yang Li, Shitao Li, Gaojun Luo, San Ling

原文摘要

Quantum locally recoverable codes with $(r,δ)$-locality ($(r,δ)$-qLRCs) are quantum error-correcting codes in which $δ-1$ erased qudits can be recovered by accessing only $r$ other qudits. The minimum locality of a qLRC is the least $r$ such that it has $(r,δ)$-locality. This paper aims to construct optimal impure $(r,δ)$-qLRCs with disjoint recovering sets and explicit minimum localities, where the optimality is measured by the Singleton-like bound of Guruswami, Kshirsagar, and Trivedi (GKT-optimal). Such qLRCs have particular advantages in recovering as many quantum erasures as possible while operating in parallel and with minimal qudit access overhead. We first characterize the recovering sets and the minimum locality of any CSS code by local relative distances, which are relative minimum distances of punctured codes with respect to shortened codes. We then construct impure qLRCs as CSS codes from nested pairs of generalized monomial-Cartesian codes. Grouping the monomials by their degree in the last variable, we reduce both their minimum distances and their local recovery to nested pairs of codes in the remaining variables, and the coordinates with the same values of these variables form disjoint recovering sets. In this way, we obtain four infinite families of GKT-optimal impure $(r,δ)$-qLRCs with disjoint recovering sets, and determine their minimum localities and the recovering sets attaining them for every admissible $δ$, except for a partial range of $δ$ in one family. These minimum localities need not be monotone in $δ$. In particular, two of the families contain GKT-optimal impure subfamilies of CSS codes CSS$(\mathcal{C},\mathcal{C})$ with $(r,δ)$-locality for some $r<n-δ+1$ at values of $δ$ for which $\mathcal{C}$ has no classical $(r',δ)$-locality for any $0<r'<n$, where $n$ is the length of $\mathcal{C}$.

cs.IT

版本更新 · v2arXiv:2603.24427

Continuous-time learning of probability distributions: a case study in a digital trial of young children with type 1 diabetes

Antonio Álvarez-López, Marcos Matabuena

原文摘要

Understanding how biomarker distributions evolve over time is a central challenge in digital health and chronic disease monitoring. For example, in diabetes, changes in the distribution of glucose measurements can reveal patterns of disease progression and treatment response that conventional summary measures miss. Motivated by a 13-week clinical trial comparing the closed-loop insulin delivery system t:slim X2 with standard therapy in children with type 1 diabetes, we propose a probabilistic framework to model the continuous-time evolution of time-indexed distributions using continuous glucose monitoring (CGM) data collected every five minutes. We represent the glucose distribution as a Gaussian mixture, with time-varying mixture weights governed by a neural ODE. We estimate the model parameters using a distribution-matching criterion based on the maximum mean discrepancy. The resulting framework is interpretable, computationally efficient, and sensitive to subtle temporal distributional changes. In the CGM trial data analyzed here, the method detects treatment-related improvements in glucose dynamics that are difficult to capture with traditional analytical approaches.

stat.ML · cs.LG

新论文 · v1arXiv:2611.25854

Learning-Based System Identification for DVL-Constrained AUV Localization

Mingxi Zhou, Emir Cem Gezer, Lin Zhao, William McConnell, Chengzhi Yuan

原文摘要

Nowadays, most AUVs utilize Doppler velocity logs (DVLs) and inertial measurement units (IMUs) to achieve satisfactory dead-reckoning results. However, the DVL bottom lock may not be available in some scenarios, such as moving over a soft bottom or traveling further away from the seafloor beyond the DVL's range. As a result, the localization system solely reliant on the IMU data will experience significant drift. The relatively high cost of DVLs also inhibits multirobot or swarm operations. In this article, we present the results of using fully connected feedforward neural networks to predict vehicle velocity (surge, sway, and heave) for AUV localization. Approximation functions for the vehicle velocities are derived based on the generalized six degree-of-freedom dynamic equation and kinematics, proving that they are suitable for feedforward neural networks with measurable inputs. The neural networks are trained using field data collected using an AUV. An ablation study was conducted to investigate the impacts of hyperparameters, and the most effective models were selected using the Akaike information criterion. The selected neural network models are tested on three new sets of mission data collected on two different days. The quantitative results show that localization using vehicle velocities predicted by neural networks yields lower averaged root-mean-square distance errors and standard deviation compared to the DVL-aided dead-reckoning method. However, the extent of this improvement was found to depend on ocean current conditions. The promising results show that lightweight fully connected feedforward neural networks offer an attractive approach for inertial-based underwater localization when DVLs or their measurements are not available.

cs.RO

版本更新 · v2arXiv:2611.00686

Generative Forecasting from Coarse Observations: Does Cascading Help?

Alexandra Spitzer, Giovanni Abati, Lorenzo Baldassari, Till Muser, Cheng Shi, Ivan Dokmanić

原文摘要

We study probabilistic forecasting of dynamical systems such as fluids and waves from coarse observations. Since the unseen fine scales influence the dynamics, the same coarse initial state can evolve into a whole distribution of futures. Generative forecasters may produce visually plausible samples, but it is hard to tell whether they sample the right conditional distribution, since trajectory data contain a single future for each initial state. We address this problem by generating reference data in a hierarchical way: for each coarse initial state, we generate many different fine-scale components and numerically evolve the full states. Using these conditional ensembles, we show that standard diffusion models and stochastic interpolants can make substantial distributional errors when trained over the full forecast horizon with finite data and model capacity, despite being exact in principle. We then show that cascading shorter forecasts greatly reduces these errors for the turbulent flows and waves we study, but we also find and characterize cases in which it hurts. Since the coarse process is not Markov, cascading discards information about the unresolved fine scales even with exact transitions. To understand this, we propose a solvable analytic model which shows that the optimal cascade depth decreases as the coupling between the unresolved and observed scales gets stronger, and increases as the dependence of the forecast on the observation becomes more oscillatory. Consistent with our experiments, cascading is detrimental when the direct forecast is "easy" to learn, for example Gaussian as in the passive scalar case. We release our datasets as a first benchmark for probabilistic forecasting with ground-truth conditional ensembles.

cs.LG

新论文 · v1arXiv:2611.25853

TableRoute: Capability-Aware Routing for Scalable Table Question Answering

Yuxiang Wang, Junhao Gan, Jianzhong Qi

原文摘要

Table question answering with large language models remains difficult to scale: tables range from compact to million-token sizes, while questions require precise grounding of columns, values, and computations. Existing methods often specialize in a fixed inference strategy, which is brittle across different table scales and model capabilities. Full-table reading is efficient for compact inputs but suffers from irrelevant context. Table reduction helps long inputs but depends on reliable evidence localization, while execution is required when tables exceed the model context window and remains sensitive to schema and value grounding errors. We propose TableRoute, a capability-aware routing framework with evidence-conditioned inference for scalable TableQA. TableRoute first generates validated question--table evidence and identifies the required operations, then routes each instance among these three strategies according to serialized table length and calibrated estimates of how reliably the base model handles these operations. The validated evidence supports the selected route by grounding columns, values, and conditions. Across WikiTQ, WTQ-L, BirdQA, and TableBench, TableRoute consistently improves accuracy through complementary routing and evidence conditioning across table scales, while reducing online token cost. Code will be made available upon publication.

cs.CL

新论文 · v1arXiv:2611.25852

Interlacing Committees with Edge-Expansion

Gili Bielous, Reshef Meir

原文摘要

Approval-based committee selection has recently been studied through the lens of \emph{interlacing}, a framework for studying polarization in approval elections, in which committees induce a similarity graph over voters. The quality of committees is then evaluated through the connectivity of the resulting graph. We revisit the connectivity measures used in this framework and propose \emph{edge-expansion}, a classical graph-theoretic measure with well-established connections to polarization in networks, as the optimization objective. We provide a theoretical foundation for this proposal by relating interlacing to classical models of information diffusion and by comparing edge-expansion with previously proposed connectivity objectives, identifying qualitative differences in the structural properties they capture. Finally, we study the computational complexity of optimizing edge-expansion, showing that hardness persists even for voter-interval preferences, while the problem is polynomial-time solvable for every fixed committee size, with an improved dependence on the committee size for voter-candidate-interval preferences.

cs.MA

版本更新 · v3arXiv:2505.04964

CAG-VLM: Fine-Tuning of a Large-Scale Model to Recognize Angiographic Images for Next-Generation Diagnostic Systems

Yuto Nakamura, Satoshi Kodera, Haruki Settai, Hiroki Shinohara, Masatsugu Tamura, Tomohiro Noguchi, Tatsuki Furusawa, Ryo Takizawa 等 10 位作者

原文摘要

Coronary angiography (CAG) is the gold-standard imaging modality for evaluating coronary artery disease, but its interpretation and subsequent treatment planning rely heavily on expert cardiologists. To enable AI-based decision support, we introduce a two-stage, physician-curated pipeline and a bilingual (Japanese/English) CAG image-report dataset. First, we sample 14,686 frames from 539 exams and annotate them for key-frame detection and left/right laterality; a ConvNeXt-Base CNN trained on this data achieves 0.96 F1 on laterality classification, even on low-contrast frames. Second, we apply the CNN to 243 independent exams, extract 1,114 key frames, and pair each with its pre-procedure report and expert-validated diagnostic and treatment summary, yielding a parallel corpus. We then fine-tune three open-source VLMs (PaliGemma2, Gemma3, and ConceptCLIP-enhanced Gemma3) via LoRA and evaluate them using VLScore and cardiologist review. Although PaliGemma2 w/LoRA attains the highest VLScore, Gemma3 w/LoRA achieves the top clinician rating (mean 7.20/10); we designate this best-performing model as CAG-VLM. These results demonstrate that specialized, fine-tuned VLMs can effectively assist cardiologists in generating clinical reports and treatment recommendations from CAG images.

cs.CV

新论文 · v1arXiv:2611.25849

EgoWitness: Benchmarking Instance-Level Spatial Memory and Reasoning in Egocentric Video Streams

Zhengru Fang, Zehao Liu, Mingyu Tsoi, Yaoyu Zhu, Junlong Li, Fengmao Lv, Sijie Cheng, Yuanliang Sun 等 10 位作者

原文摘要

Recalling where and when an object was last seen and who saw it requires recognizing it across days and people. Existing long-video benchmarks focus on events, object categories, or relative locations, leaving instance-level spatial memory largely untested in multimodal models. To address this gap, we present EgoWitness, a benchmark for instance-level spatial memory and reasoning in egocentric video. Its 8,453 human-verified questions span approximately 324 hours of week-long and short-horizon recordings and test seven abilities, from recognizing an object or room to recalling who saw it and reasoning about room visits. Over one million reviewed masks of 34 registered landmarks and 21 movable-object labels are linked through room timelines. Among twelve direct vision-language models (VLMs), open-weight models reach at most 46.97% accuracy, and none exceeds 35% on identifying who saw an object last. We therefore propose EgoClue, a spatial-memory agent that retrieves candidate intervals, verifies instance identity, and computes spatial and temporal relations with five tools. Two-stage training with supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO) raises its 4B controller from 23.64% to 72.65% accuracy on 1,320 held-out questions. Evidence-trace synthesis matches the best direct VLM at 74.85% on week-long questions and achieves 80.56% in environments excluded from training.

cs.CV

新论文 · v1arXiv:2611.25848

Interpretable Prototype-Conditioned Adaptation for Weather Forecasting

Zhixuan Li, Hao Jiang, Chen Hui, Kai Geng, Weisi Lin

原文摘要

Data-driven weather forecasting requires distinguishing local states and their temporal evolution. However, learned patterns remain implicit in neural forecasters, making their influence difficult to interpret. Failing to distinguish local evolution can lead to similar forecasts for locations that develop differently, increasing prediction error. This motivates explicit references for distinguishing and interpreting local patterns. Prototypes provide explicit pattern references, but sample- or task-level conditioning leaves open how to adapt to distinct locations and variable groups. To deal with this problem, we design ProtoCast, which uses group-specific prototype banks to adapt local features while accounting for cross-variable dependencies. Specifically, (1) we combine each group's historical features and temporal change with forecast context from cross-variable attention for evolution-aware prototype matching. (2) Sparse mixtures of prototype-specific low-rank operators tailor corrections to each group and location. A gated attention block then exchanges the corrected features across groups before prediction. Each group's correction decomposes into explicit prototype contributions before this fusion. (3) We jointly train the forecasting network and operators from scratch and refit each group's prototypes before every training epoch. Slot tracking preserves prototype-operator associations, making each prototype's contribution to local corrections more interpretable. Experiments on GlobalForecast and TyphoonTracking show improved forecasting accuracy, while prototype deletion links local feature contributions to changes in prediction error. This work advances the development of more accurate and interpretable data-driven weather forecasting models. Project page: https://zhixuanli.github.io/projects/li2026protocast

cs.LG · cs.CV

版本更新 · v2arXiv:2610.10592

Wieszcz-XIX: A 3.1-Billion-Word Corpus of Pre-1918 Polish and Temporally Bounded Language Models Trained From Scratch

Szymon Kocur

原文摘要

Historical Polish is well documented as a language but annotated in machine-readable form only to about a million words for the period this paper covers; the rest sits behind optical character recognition of variable quality. We present Wieszcz-XIX, a corpus of 6.75 billion tokens (about 3.1 billion words) in 294,369 documents, most of them periodical issues, of Polish published from 1800 to 1918, assembled from Wolne Lektury and the Internet Archive by a pipeline that filters, deduplicates, audits for post-1918 leakage and splits at the document level. It is over three orders of magnitude larger than the annotated corpus of the same period, and we quantify its defects: recognition corruption against a false-positive floor, near-identical duplication, which is removed, and post-1918 leakage, which is excluded from the training corpus itself down to a known residue of 0.04 to 0.38% of its bytes, found in the transcribed source, so the published corpus is the trained one document for document. On a hand-corrected sample the character error rate is 0.68% where the text is legible, and 45% of the sampled passages cannot be corrected. On it we train a ladder of decoder-only models from 47M to 349M parameters from scratch, and measure their temporal boundedness. Against two modern Polish base models, one far larger, the 349M shows a crossover, as does the 107M against the comparator of its size: post-1918 vocabulary costs them about 3.1 bits per byte more than period vocabulary, a gap the comparators do not show, and period vocabulary costs them fewer bits than it costs the comparators. Shown period text, the models keep its spelling and the comparators only partly. Adding parameters gains about twice as much as a second pass over the data. We release the corpus, code and weights. Content warning: the models reproduce period prejudice, including antisemitic statements.

cs.CL · cs.DL

新论文 · v1arXiv:2611.25847

TrajCal: Trajectory-Calibrated Time Schedules for Few-Step Flow Sampling

Junchi Zhang, Zihan Gu, Jiangcan Jia, Lulu Zhao, Wenbo Su, Hua Zhang

原文摘要

Flow matching enables high-quality image and video generation, but its iterative sampling process requires repeated evaluations of large neural networks. Existing few-step methods can distort the output and reduce its agreement with a many-step reference. We propose TrajCal, a training-free calibration framework for few-step sampling. Based on scatter search, a well-established metaheuristic for difficult optimization problems, TrajCal starts with diverse seed schedules, deforms and combines these candidates, and selects one fixed schedule by final output fidelity for reuse on new inputs. For both image and video generation, TrajCal achieves higher reference fidelity than all compared baselines at low evaluation budgets while maintaining competitive visual quality. The search is effective even with randomly generated seeds, while trajectory-based seeds further improve fidelity. For cross-style generalization, photography-calibrated schedules also improve fidelity on unseen image styles.

cs.CV

新论文 · v1arXiv:2611.25845

LINEAGE: A Benchmark for Empirical Lineage Retrieval in AI Scientific Literature

Pangjing Wu, Xu Yuan, Wenqi Fan, Qing Li

原文摘要

Current research agents are moving beyond academic search toward literature synthesis and experiment planning. These workflows increasingly rely on literature retrieval systems, which serve as standalone systems or as components of research agents. However, existing scientific literature retrieval benchmarks mainly evaluate these methods using topical relevance or citation links as supervision, which cannot capture which prior methods are used for empirical comparison. To address this gap, we introduce Lineage, a benchmark for empirical lineage retrieval in AI scientific literature. An empirical lineage relation links a prior paper to a source paper that explicitly uses its method as an empirical comparison baseline. These relations form lineage graphs whose multi-hop neighborhoods provide retrieval ground truth for structural evaluation. To construct these graphs at scale, we develop Lineage-Constructor, which converts heterogeneous paper records and full-paper empirical baseline evidence into calibrated lineage graphs. Built from 121,048 papers across 17 top AI venues from 2018 to 2025, Lineage contains 3,950 graph components, 67,502 paper nodes, 164,948 lineage edges, and 80,000 multi-facet retrieval queries. Experiments with 19 representative retrieval systems show that empirical lineage retrieval remains challenging. Performance drops sharply with lineage distance, while structural metrics reveal that isolated retrieval hits often fail to recover closely connected empirical comparison neighborhoods.

cs.IR · cs.AI

版本更新 · v3arXiv:2608.22294

Semantic Role-Structured World Model for Physical Interaction Planning

Juntao Cheng, Yijun Shen, Qunchao Jin, Xiansheng Chen, Wei Liu, Rui Qian, Jingyi Peng, Yunsong Zhou 等 9 位作者

原文摘要

World models for robotic manipulation must capture how actions transform a scene in ways that are meaningful for task execution. Existing object-centric approaches preserve the transient identity of entities, while leaving the evolving task structure that gives those dynamics meaning largely implicit. We introduce SemRole, a world model that represents scene evolution through semantic roles: a task-conditioned structure explicitly encoding the functional status of scene entities, their task-relevant relations, and the current phase of execution. Conditioned on language, visual observations, and proprioception, SemRole uses this structured state to govern action generation and action-conditioned dynamics, predicting interaction events, relation satisfaction, and phase progression. At inference time, these predictions guide phase-aware action selection and local repair of invalid rollouts. We instantiate SemRole on both lightweight LeWM and pretrained Cosmos backbones, improving closed-loop success from 45.5% to 83.0% on LIBERO, 42.9% to 55.2% on RoboCasa, and 55.0% to 81.0% on real-robot Galbot transfer. Across backbones and domains, explicitly modeling task-semantic structure yields substantially stronger world-model-based manipulation.

cs.RO

新论文 · v1arXiv:2611.25840

Harnessing Medical Visual-Language Models with Agentic Evidence Management

Qiming Huang, Hanze Dong, Jipeng Zhang, Mike Zheng Shou, Xinxing Xu, Furu Wei

原文摘要

Medical vision-language agents increasingly rely on segmentation tools, and recent work has focused on acquiring the right mask, often treating a good mask as the end point of tool use. But do these agents use the resulting evidence effectively? To answer this question, we need a benchmark that separates evidence use from segmentation quality, spans multiple medical image types, and allows the same mask to be presented in different ways. Existing medical VQA datasets do not jointly provide these properties. We therefore construct Mask-Evidence MedVQA, a 2.5K-example benchmark spanning four image types, where each question is paired with an answer-relevant ground-truth mask and multiple possible renderings. Our study reveals two limitations of direct tool use. First, evidence presentation matters. Different renderings of the same fixed mask can substantially change model behavior. Second, masks introduce an evidence trade-off. Highlighting a relevant region can expose diagnostic evidence the model previously missed, but the added visual cue can also induce incorrect associations. As a result, even a ground-truth mask can correct some answers while turning initially correct ones into incorrect ones. Together, these findings show that acquiring the right mask is not enough. An agent must also decide how to present the evidence and whether it should change the answer. We therefore introduce TUNE (Tool Use is Not Enough), an evidence-management harness that explicitly manages what evidence to acquire, how to present it, and whether to accept the resulting answer update. Across three public medical VQA benchmarks, evidence management improves the fine-tuned Qwen3.5-9B backbone's mean accuracy by 4.0 percentage points. Full TUNE improves over direct tool use by 9.1 points on average across seven benchmarks and achieves competitive results against existing medical VLMs and agents.

cs.CV · cs.AI

新论文 · v1arXiv:2611.25839

Self-Replicating Language Models

Arend Hintze

原文摘要

Whether large language models (LLMs) and LLM-based agents are a form of life is a scientifically interesting question. Under the Darwinian definition of life, it is also a practical one, because a system capable of Darwinian Evolution can adapt. We therefore test several modes in which a language model can reproduce itself: by emitting a genome that trains its offspring, by copying and mutating its own weights, and by teaching a child from its own samples. In principle, small language models can replicate and evolve: selection improves them, and they adapt to narrow niches. Selection also lowers mutation rates, and a language model, unlike any organism, can lower its own to zero, which may limit its evolution. Flagship LLMs can write such replicators, but for a large LLM to replicate by teaching its offspring remains, for now, technically and economically infeasible.

cs.AI

新论文 · v1arXiv:2611.25835

PSM-WSM: Predictively Sufficient Multimodal World Social Modeling

Chunlei Meng, Zhenyu Yu, Pengbin Feng, Jacqueline J. Pang, Jiabin Luo, Chih-Ting Liao, Zhaolu Kang, Chun Ouyang 等 9 位作者

原文摘要

Learning latent states that support future evolution remains an open problem for multimodal interactive systems. Existing approaches typically optimize representations for current semantic understanding or direct prediction, without explicitly defining what information a latent state should preserve to remain predictive of future dynamics. In this work, we investigate this problem through the principle of predictive sufficiency, which defines a predictive state as a latent variable that retains future-relevant information from historical multimodal observations while reducing redundant dependence on the observed history. Based on this principle, we propose Predictively Sufficient Multimodal Social World Modeling (PSM-WM), a framework that jointly learns predictive latent states and their stochastic temporal evolution through probabilistic state inference, event-conditioned transitions, and multi-step consistency optimization. Beyond forecasting performance, we introduce state-centric evaluations to measure whether learned representations satisfy predictive state properties, including predictive closure, residual history gain, and future semantic alignment. Experiments on Hi-EF, IEMOCAP, MELD, and Ego4D show that PSM-WM improves future multimodal interaction modeling and learns latent states with stronger predictive properties and robustness. These results indicate that predictive sufficiency provides an effective principle for constructing future-oriented multimodal states and highlight the importance of state construction in multimodal world modeling.

cs.MM

版本更新 · v5arXiv:2610.19246

Refined Topology-Pass Scheduling: Damped Trust-Region Compensation for Dynamic k-NN Graphs in SE(3)-Equivariant Structure Prediction

Jiankai Huang, Hanwen Shen

原文摘要

Equivariant structure-prediction pipelines alternate coordinate updates with the construction of a k-nearest-neighbour (kNN) graph whose membership is a discrete, piecewise-constant function of the coordinates. Rebuilding that graph at every block boundary (PB) is the expensive part of an epoch, whereas reusing one frozen graph (F) is fast but leaves a topology gap. We study a two-pass design in which the cheap pass records a small amount of state (margins, kNN radii, measured per-boundary displacements) and a second pass compensates conservatively. We formulate the choice of the graph-construction basis as a damped fixed-point iteration whose unique fixed point is exactly PB, and derive (i) exact non-flip certificates with the two-sided constant $m_i/4$ and one-sided constant $m_i/2$, (ii) the minimum-variance (BLUE) replay gain $γ^*$ and the resulting feasibility crossover, (iii) an $\ell_\infty$ trust-region projection that preserves convergence, and (iv) an analytic near-tie spacing law $E[m] \approx r_k/(3k)$ that matches measurements to 1\%. Three compensation levels are implemented in C++ (L1: frozen set with live geometry; L2: margin-gated local repair; L3: damped, projected basis) and validated under a matched-conditions protocol (identical seed, MSA depth and trajectory family, only the topology mode differs). L2 reproduces the PB edge set exactly ($D = 0.0000$ at every boundary), L1 stays within 0.6% and L3 interpolates monotonically: L2 reproduces the PB edge set exactly when the trust region is removed, and L3 interpolates between the frozen and the PB topology. The recorded trajectory itself yields a measured $1.5\times$ Pass-1 and $1.76\times$ epoch speed-up with no measurable loss gap. An open-source implementation of the runtime and of all the experiments reported here is available at https://github.com/AlbertHuang96/ProteinPredictMachineLearning.

cs.LG · cs.AI

版本更新 · v3arXiv:2611.22639

Hierarchical Control of Antiphase Synchronization in Biomorphic Spiking Neural Architectures for Locomotor Control

Artem Guba, Nikolai Kovalev, Alexey Zharinov, Vyacheslav Demin, Viktor B. Kazantsev

原文摘要

Central pattern generators (CPGs) provide a biologically inspired foundation for adaptive locomotor control, yet systematically characterizing their dynamic regimes in spiking neural networks remains challenging. This study introduces a framework for identifying and stabilizing antiphase synchronization in a biomorphic spiking CPG model embedded in a closed neuromuscular loop. A composite synchronization metric is developed, integrating phase coherence, amplitude balance, and temporal stability. Its performance is evaluated across systematically varied stimulation waveforms and synaptic configurations. Statistical analysis reveals that stimulation type dominates regime determination, with tonic drive yielding the most robust antiphase synchronization. Periodic inputs induce nonlinear saturation and context-dependent network routing. Although synaptic weights alone show limited discriminative power, they significantly modulate robustness within a given stimulation context, with reciprocal Ia interneuron connections emerging as key stabilizers. The proposed composite metric achieves robust antiphase detection, outperforming traditional phase-only measures while preserving biophysical interpretability. These findings establish a statistically validated, hierarchical modeling approach in which external drive defines the operating regime and targeted synaptic configurations fine-tune dynamic stability, providing actionable guidelines for closed-loop neuromechanical controllers and adaptive neuroprosthetic interfaces.

q-bio.NC · cs.AI

新论文 · v1arXiv:2611.25832

RAVEN: A Self-RAG Based Agent for Threat Classification and Response

Pironti Francesco Aurelio, Arena Luigi, Blefari Francesco, Lupinacci Matteo, Romeo Francesco, Furfaro Angelo

原文摘要

Security event analysis in large enterprises can be overwhelming, as each alert typically requires manual in- spection by human operators. This process is time-consuming and prone to inconsistencies, especially in high-volume Secu- rity Operations Center (SOC) environments. In this work- in-progress paper, we present RAVEN: Retrieval-Augmented Vulnerability & Event aNalysis, an agentic framework designed to support SOC analysts by automating the initial stages of alert investigation. The proposed system leverages a Self- Reflective Retrieval-Augmented Generation-based approach to provide an initial classification of security events and enables interactive refinement through analyst-in-the-loop queries. Un- like traditional approaches that rely on generic mitigation strategies, RAVEN generates context-aware remediation sug- gestions by incorporating knowledge of the target enterprise infrastructure. We implement an early-stage prototype and report preliminary results on three balanced test sets of 400 curated web payloads each (SQLi, XSS, SSTI) and a qualitative case study of mitigation generation; an evaluation on IDS- generated alerts and a user study are ongoing and outlined as remaining work.

cs.CR

版本更新 · v2arXiv:2609.16546

GPUThor: Amplifying Rowhammer Attacks via Non-Uniform Patterns to Exploit ECC-Protected GPUs

Chris S. Lin, Joyce Qu, Aditya Rajeev, Gururaj Saileshwar

原文摘要

GDDR memory in GPUs is vulnerable to Rowhammer attacks, where rapid memory accesses induce bit flips in adjacent cells, enabling data tampering and privilege escalation. However, prior GPU Rowhammer attacks trigger only tens to hundreds of bit flips, orders of magnitude fewer than CPU attacks, severely limiting their practical impact. This gap stems from the reliance of existing GPU Rowhammer attacks on uniform hammering patterns that activate aggressor and decoy rows equally, which results in low hammering intensity for aggressor rows. We present GPUThor, a high-intensity Rowhammer attack on NVIDIA GPUs leveraging non-uniform hammering. GPUThor reverse engineers GPU memory-access coalescing behavior to enable non-uniform hammering patterns on GPUs, that activate aggressor rows more intensely than decoy rows. Additionally, by identifying refresh instances when in-DRAM mitigations are applied, it constructs longer attack patterns that escape mitigation across refresh intervals, further increasing hammering intensity. Together, these techniques yield 500X to 23,500X more bit flips than prior GPU Rowhammer attacks, across several NVIDIA GPUs (A4000, A4500, A5000, A6000), reaching bit flip rates close to state-of-the-art CPU Rowhammer attacks. GPUThor also enables the first Rowhammer exploits on ECC-protected GPUs, inducing uncorrectable double and triple bit flips, making denial-of-service and privilege-escalation attacks practical even on GPUs with ECC enabled.

cs.CR

新论文 · v1arXiv:2611.25831

Beyond Alignment: Whitening Geometry, Profile Synthesis, and Heterogeneous Ensembles for Cross-Environment WiFi Action Recognition

Lei Wang

原文摘要

Cross-environment deployment is the central obstacle to practical WiFi sensing: a pipeline reaching 0.89+ top-1 in its training environments collapses to 0.408 under Leave-One-Domain-Out (LODO) evaluation with only K-shot labels. A ceiling analysis shows the collapse is not an alignment problem: the action signal is strong across seen environments (about 0.89 K10), environment identity is decodable from the encoder (0.965), and the failure concentrates in the encoder's extrapolation to the novel environment. A pre-registered falsification chain confirms that "align the domains" methods--training-time alignment, readout-level prototype alignment, adversarial confusion, and covariance/subspace alignment (CORAL -3.8pp, subspace -8.3pp)--degrade performance, and that source-domain statistics anti-correlate with cross-environment contribution. The effective levers lie elsewhere; we combine them into Dec-H: (i) closed-form source within-class whitening, provably re-geometrizing the readout's L2 regularization (+8.6pp, zero cost); (ii) profile-mixing augmentation (+1.0pp); (iii) few-shot adaptation with alternating whitening refit (+1.5pp; fixed preconditioners are 20/20 negative); (iv) heterogeneous decision ensembles (+1.7pp); homogeneous scaling and source-weighted stacking are saturated or harmful. Dec-H reaches 0.560 K10 (0.812 K40), +15.2pp over the baseline, consistent across all four environments and seeds; all thirteen configurations were pre-registered with frozen criteria.

cs.LG · eess.SP

版本更新 · v2arXiv:2611.21799

Controllable Chains of Thought Can Reveal Less, Even Without Evasion Training

Zheng-Xin Yong, Benjamin Arnav, Miles Wang, Tomek Korbak, Olivia Watkins

原文摘要

Chain-of-thought (CoT) monitoring is a promising tool for detecting hidden goals, off-task behavior, and other failures in AI agents. One risk is that models could deliberately shape their reasoning to make it opaque to monitors. We stress-test this risk through persona-control training: supervised fine-tuning on mathematical reasoning traces rewritten in the voices of diverse personas. This induces a broader ability to control CoT, even though training contains no monitor-evasion examples. In covert side-task evaluations, traces in which a gpt-oss-20b model organism completes the hidden side task contain less monitorable evidence than the base model's. Averaged across evaluations and CoT lengths, weighted by side-task-success counts, the CoT monitorability metric g-mean$^2$ falls from 0.94 for the base model to 0.68 for the model organism without a hiding instruction, and to 0.22 when explicitly instructed to hide. In intervention evaluations, explicit hiding often drives monitorability near zero. The effect extends to a Kimi-K2.6 model organism in agentic environments, though hidden objectives requiring extended reasoning or longer trajectories remain more monitorable than simple single-step side tasks. We further discover that targeted transparency prompts can substantially recover monitorability, sometimes surpassing the base model. These model organisms provide a testbed for studying monitorability failures and potential mitigations. Our code can be found on https://github.com/yongzx/cot-control-model-organism.

cs.AI · cs.LG

新论文 · v1arXiv:2611.25826

A Sparse Low-Rank Biclique Decomposition for Graphs

Antoine Vialle, Sergei Gerasimov, Aref Einizade, Antonio Ortega, Fragkiskos D. Malliaros, Jhony H. Giraldo

原文摘要

Low-rank plus sparse decompositions, such as robust PCA, typically model the low-rank component as dense and the sparse component as corruption. This viewpoint is poorly suited to graph adjacency matrices, where sparsity is structural and dense low-rank factors destroy the combinatorial and computational properties of the graph. We propose signed biclique decomposition (SBD), a sparse low-rank decomposition for integer-weighted graphs that represents the structured component as a sum of sparse signed binary rank-one factors, together with a sparse signed residual. The resulting representation remains discrete, interpretable, sparse, and compatible with standard linear algebra. We formulate fixed-rank and greedy SBD objectives and derive a gain-based greedy algorithm with theoretical guarantees. Empirically, SBD recovers block structure in noisy graphs, while its factorized representation enables efficient sparse matrix-matrix multiplication (SpMM). On real graphs with strong shared-neighborhood structure, SBD achieves compression factors above $10$ and SpMM speedups up to $6\times$ over cuSPARSE, suggesting a practical alternative to dense low-rank decompositions for graphs. Code is available at https://github.com/IAntoineV/sbd.

cs.LG

新论文 · v1arXiv:2611.25824

Fusing Multimodal CNN Features with Structurally-Masked Transformers for Hierarchical Sub-Pixel Tree Species Mapping

Mattia Ferrari, Lorenzo Bruzzone

原文摘要

Precise forest species mapping using spaceborne hyperspectral data is challenging due to high spectral similarity and the inherent sub-pixel heterogeneity of natural forest stands. While traditional deep learning models exploit spatial context through convolutional architectures, these approaches can introduce noise in fragmented or mixed-species forests. In this paper, we propose a hierarchy-aware, multimodal deep learning framework to resolve sub-pixel forest composition by integrating PRISMA (PRecursore IperSpettrale della Missione Applicativa) hyperspectral imagery, high-resolution panchromatic data, and digital elevation models. The architecture features a multi-branch encoder combining an EfficientNetV2-based one-dimensional convolutional neural network (CNN) for spectral signatures and twin two-dimensional CNNs for spatial and topographic features, coupled with a novel structurally-masked Transformer. By using a binary transition tensor to condition the Transformer slot embeddings and constrain inference, the model enforces taxonomic consistency between broad forest categories and fine-grained species distributions. The framework is validated across three diverse Italian biogeographical regions (the Alps and the Central and Southern Apennines) using field-surveyed plots. The proposed method consistently outperforms state-of-the-art multi-label and ranking-based paradigms, particularly in identifying subordinate species and their relative dominance. The bidirectional learning mechanism allows sub-pixel compositional features to refine macroscopic forest type predictions, while the integration of topographic priors effectively resolves spectral ambiguities in complex terrains. This knowledge-driven method supports regional-scale biodiversity monitoring and forest characterization within the evaluated PRISMA scenarios.

cs.CV

新论文 · v1arXiv:2611.25822

Limits of local curvature for constructing search subspaces in segmentation models

Maxim Bolshim, Alexander Kugaevskikh

原文摘要

We examine how well subspaces constructed from local curvature represent the change in the gradient update field after a finite step in semantic segmentation models. The local response is the action of the curvature operator on the actual input displacement. In a study of 50 images and two models, we measure the ratio of the loss increase attained in a fixed subspace to that attained in the full space. This ratio decreases as the search radius increases. On ten images from the same sample, we compare the one-dimensional subspace spanned by the initial gradient with two rank-16 subspaces for both models. The two equal-rank subspaces are assessed at the same initial states, using the same actual steps and measured vectors. Adding generalised Gauss-Newton directions to the initial gradient reduces the mean relative error in representing the field change from 0.922 to 0.874 for the first model and from 0.819 to 0.763 for the second. At the same rank, including the response direction for the initial gradient nearly eliminates the error in representing the local response. The mean error in representing the field change does not decrease. A decomposition with fixed pixel weights shows that the finite-step remainder is comparable in norm to the field change and makes the largest diagonal contribution to the squared projection error of the field after the step. Evaluation of a search subspace should measure how accurately it represents the field change and the local response.

cs.CV · cs.LG

新论文 · v1arXiv:2611.25821

Signed but Unbound: The Authority Gap Between Agent Commerce Mandates and Crypto Wallet Signatures

Shengwei You, Aditya Joshi, Andrey Kuehlkamp, Jarek Nabrzyski

原文摘要

When an artificial intelligence agent pays with a stablecoin, the commerce protocol can specify the merchant, order, amount, and facilitator. The token contract enforces the fields the wallet signs. We call the difference the authority gap, define nine constraint classes, and derive them for 12 authorization configurations, including Coinbase's deployed Commerce Payments Protocol. All 31,800 trials on local copies of deployed contract code match the derivation. A controlled comparison shows that committing the terms in the signed nonce binds the order. When the facilitator submits a payment to a merchant account without code, binding the facilitator requires a separate recipient contract. Of 26,653 sampled USD Coin authorizations redeemed on Base, 98.7% followed a path that any holder of the signature could redeem. We prove sufficient conditions for a recipient contract to enforce the signed mandate under a trusted signer and show which recipient a wallet displays. MandateLock, a minimal implementation for immediate payments, rejects all 10,800 modified mandates on deployed token code and accepts signatures from two independent signers. It uses 25% more gas than a transfer that only the recipient can submit.

cs.CR