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

+1112新增论文
+875其中 AI 相关
+133已有论文修订
+95新增 AI 导读

arXiv 最近完成扫描:2026-11-20 00:30 · 全站最近入库:2026-11-19 19:09各信源进度 →

统计口径与数据新鲜度

统计 2026-11-19 00:31 — 2026-11-20 00:31(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:2026-11-20 00:30最近发现新增版本:2026-11-19 19:09库内论文最新更新时间:2026-11-19 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。

27,982 已收录论文23,249 AI 相关8,559 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 20,663 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-20 00:28(北京时间)
版本更新 · v2arXiv:2602.08693

Reasoning under uncertainty in humans and language models

Gonçalo Guiomar, Elia Torre, Victoria Shavina, Pehuen Moure, Mario Giulianelli, Shih-Chii Liu, Valerio Mante

原文摘要

Do language models make decisions under uncertainty like humans do? And if so, what role does extended reasoning play in the underlying decision process? We answer this question by introducing an active probabilistic reasoning task that cleanly separates sampling (actively acquiring evidence) from inference (integrating evidence towards a decision). Benchmarking humans and a broad set of contemporary LLMs against an optimal reference policy reveals a consistent pattern: extended reasoning is the key determinant of strong performance, driving large gains in inference, while yielding only modest improvements in active sampling. To explain these differences, we fit a behavioral model that captures systematic deviations from optimal Bayesian behavior through interpretable parameter families, placing humans and models in a shared low-dimensional cognitive space. The resulting fits show how reasoning shifts models toward human-like regimes of evidence accumulation and belief-to-choice mapping, and yield testable predictions about the latent dynamics that might drive each decision. Probing residual-stream activations of an open-weight reasoning model, we find that the geometry of internal representations tracks these predicted dynamics, linking behavior to representational correlates of the fitted latent dynamics.

cs.LG

新论文 · v1arXiv:2611.20927

LEDGER: Auditable Agent Memory Under Change

Md Aidul Islam, Md Nasir Uddin Shuvo, Muhammad Waseem, Pekka Abrahamsson

原文摘要

Agent memory systems are mostly evaluated on fixed histories, and only their final answers are scored. In real use, facts change, sources differ in reliability, and users ask the agent to forget things. We present LEDGER, a memory layer in which a language model proposes facts and explicit rules decide which value is current, which source to trust, and what to delete. Each fact keeps its source and the period in which it held. LEDGER+EP adds a search over raw conversation turns for questions the ledger does not cover, and scrubs these turns on deletion. We compare both against no memory, full-history replay, retrieval over raw transcripts, and Mem0, with the same answering model and prompt, on a synthetic multi-session coding-agent workload and a 210-question subset of LongMemEval, and we score both the evidence each memory sends to the model and the answer. On the coding workload, LEDGER+EP in fallback mode answers 94.3% of questions correctly on five development seeds and 91.1% on five held-out seeds, against 82.1% and 83.3% for full replay, the strongest baseline. When a deletion request only describes the fact, no deleted value reached our systems' evidence in any evaluated probe. On LongMemEval, LEDGER+EP in augment mode has the same observed accuracy as full replay (74.8%) with 1% of its context. The main limitation is that the coding workload is synthetic and built by us; on LongMemEval, the differences between the strongest systems are within the confidence intervals.

cs.MA · cs.PF · cs.SE

新论文 · v1arXiv:2611.20926

Advancing Cognitive Pairwise Comparisons for Human-Centric Artificial Intelligence and Data Science: A Human-in-the-Loop Framework

Kevin Kam Fung Yuen

原文摘要

Human-centric artificial intelligence and data science (HCAD) systems increasingly involve human-in-the-loop (HITL) interventions to overcome the limitations of fully automated machine learning algorithms, such as automation bias and the inability to incorporate tacit domain knowledge, and continuously improve predictive performance. Data transformations, feature selection, feature weighting, and model selection usually require laborious human involvement across iterations of artificial intelligence and data science (AI&DS) pipelines, making development a challenging and time-consuming process. The Cognitive Pairwise Comparison (CPC) framework is proposed as a human-centric methodology to elicit preferences and priorities in a structured format. To establish essential facets of human-centricity, this research advances the CPC methodology so that user preference elicitation can be systematically quantified in these challenging AI&DS tasks. To demonstrate the usability and validity of these facets, this study presents several applications within movie recommender systems using the MovieLens dataset. It exemplifies how the extended CPC framework can facilitate interactions between human judgments and data across supervised learning, dimensionality reduction, and unsupervised learning algorithms.

cs.AI · cs.HC · math.OC

新论文 · v1arXiv:2611.20924

Navital Gym: Rule-Graded Agent Environments for Indian Workflows

Daksh Bathla

原文摘要

AI agents are being deployed on Indian workflows such as payments, tax, forms, travel and customer support in Hindi and Hinglish. Public agent evaluations rarely cover these workflows, and many score with another language model. We present Navital Gym, a set of six open environments with 321 public tasks in five task families: trip planning, UPI payments and disputes, GST invoicing and returns, form filling (including a multi-document split) and customer support in Hindi, Hinglish and three experimental Indian languages. Every task is graded by named, deterministic rules that recompute the outcome from a synthetic sandbox world. No language model judges any score. We report one run per task for three models on all six environments, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 4.5, and for Gemini 3.1 Flash-Lite on two (UPI and the Support core). Opus solves 315 of 321 tasks, Sonnet 302 and Haiku 205. Trip planning and Hindi/Hinglish support separate the models most. We also describe a private held-out set that is not published, and a 7B adapter trained in the gym. Code is Apache-2.0 and task data is CC BY 4.0. The rules are a documented simplification of Indian practice, not the live systems.

cs.CL

新论文 · v1arXiv:2611.20921

Life-cycle structural reliability of RC bridge piers under corrosion in a changing climate

Giuseppe Vittorio Nava, Luca Capacci, Fabio Biondini, Lorenzo Casti

原文摘要

Aging and deterioration processes of structures and infrastructure systems can be severely affected by the long-term evolution of environmental conditions expected in a changing climate. This paper aims at investigating the interaction between climate change and environmental aggressiveness on the probabilistic life-cycle performance of concrete structures under corrosion. A simulation-based framework is proposed for life-cycle structural reliability analysis under climate change scenarios. The framework is applied to a bridge pier with hollow core cross-section exposed to chloride-induced corrosion. The results allow to quantify the impact of climate change on long-term probabilistic structural response and highlight the need for proper modeling of the environmental parameters and climate evolution over the entire lifespan of aging structures for optimal life-cycle maintenance and management of structural and infrastructural facilities.

cs.PF

新论文 · v1arXiv:2611.20916

EviBridge: Encoding Predictive Visual Evidence for Proactive Interaction

Zhichao Huang, Changheng Cai, Song Ren, Zhichen Zhao, Hongsen Liu, Chaoyang Zheng, Wenxi Wu, Yichen Long 等 12 位作者

原文摘要

Tool-handover interaction calls for preparation before a response is warranted. The model waits until interaction cues appear, then prepares while assessing whether the accumulating evidence supports a response or calls for cancellation. We introduce EviBridge, a framework that integrates predictive visual representations into a vision-language interaction model for joint state and tool prediction. A trainable evidence interface maps features from a frozen V-JEPA 2 encoder into the language decoder alongside native visual inputs, preserving their temporal context and using only observations available at each decision time. The framework has two instantiations. EviBridge-E distinguishes recent from historical evidence and encodes the timing and spatial location of each observation. EviBridge-I preserves the temporal order of pooled V-JEPA 2 features and interaction turns and generates dedicated interaction tokens. We evaluate on FuGIBench, a streaming-video benchmark for preparation, response, and cancellation in tool-handover assistance. The best-performing variants of EviBridge-E and EviBridge-I achieve Overall scores of 0.6448 and 0.5892, compared with 0.2178 and 0.5327 for their respective task-adapted native baselines. These gains span target handovers, long waits, and cancelled requests, demonstrating the value of predictive visual evidence for interaction decisions.

cs.CV

新论文 · v1arXiv:2611.20915

Multi-LLM Agentic AI at the Edge: Joint Agent Scheduling and Resource Allocation

Yuanyan Song, Kezhi Wang, Xinmian Xu, Yongmin Li

原文摘要

In this paper, an edge-enabled multi-large language model (LLM) agentic AI architecture is proposed, where one supervisor AI agent and multiple worker AI agents cooperatively execute users' requests at the edge. The supervisor AI agent decouples the user request into a series of sub-tasks and dynamically dispatches them to LLM-equipped worker AI agents via two agentic workflows, i.e., parallel and serial workflows, with a two-level retry mechanism to enhance execution reliability. To prolong the service duration of worker AI agents, the energy minimization problems are formulated by jointly considering AI agent selection and resource allocation, subject to task execution latency and quality of service (QoS) constraints. A resource allocation scheme based on alternating optimization (AO) is proposed. Specifically, for the parallel workflow, the scheme decouples the original problem into bandwidth allocation, retrieval selection and AI agent selection sub-problems, which are iteratively solved by the proposed successive convex approximation (SCA), Lagrangian relaxation (LR), and Gale-Shapley (GS) or LLM-based algorithms, respectively. For the serial workflow, the problem is similarly decoupled into service time allocation, retrieval selection and AI agent selection sub-problems; the optimal service time allocation is obtained via Karush-Kuhn-Tucker (KKT) conditions, while the latter two sub-problems are addressed by extending the proposed algorithms to the serial scenario. Simulation results show that the proposed resource allocation scheme can reduce system energy consumption compared with other benchmarks, while experiments on a multi-LLM agentic AI system based on Qwen and Llama demonstrate improved task execution success rates.

cs.AI

新论文 · v1arXiv:2611.20914

Analysis and Design of Spatially Coupled LDPC Codes for Pulsed-Jamming-Resilient Space Telemetry

Massimo Battaglioni, Mauro M. M. Costantino, Rebecca Giuliani, Franco Chiaraluce, Marco Baldi

原文摘要

Jamming is a major threat to space telemetry downlinks, which operate under severe power and bandwidth constraints. This paper develops a jamming-aware analysis and design framework for protograph-based spatially coupled low-density parity-check (SC-LDPC) codes. For pulsed jamming with jamming state information (JSI) at the receiver, we derive a two-state protograph-based extrinsic information transfer (PEXIT) recursion that tracks the messages originating from clean and jammed symbols separately and merges them at the check nodes through a mutual-information-matched Gaussian approximation. For non-interleaved pulsed jamming, we map each pulse realization, through the serialization of the lifted codeword, to a profile of jammed fractions over the variable-node types, and we prove that, under random lifting, this profile determines the local channel-state statistics. The resulting position-dependent thresholds are the objective of a constrained differential-evolution search over protographs, coupling memories, and edge spreadings. At different code rates, the designed ensembles are compared with the Consultative Committee for Space Data Systems (CCSDS) accumulate, repeat-by-4, and jagged accumulate (AR4JA) LDPC codes whose block length matches the SC-LDPC constraint length, with longer AR4JA codes whose block length is comparable with the size of the sliding window decoder, and with non-optimized SC-LDPC codes. Under pulsed jamming with random pulse positions and with JSI, the ensembles designed over random pulse realizations, decoded through sliding window decoding, reach a bit error rate of $10^{-4}$ at lower signal-to-noise ratios than all the benchmarks.

cs.IT

版本更新 · v3arXiv:2510.12546

EmpathiSEr: Validated Scales for Measuring Empathy Towards Software Engineering Practitioners and Users

Hashini Gunatilake, John Grundy, Rashina Hoda, Ingo Mueller

原文摘要

Empathy plays a critical role in software engineering (SE), for instance in situations where developers interpret non-technical users' frustration with system usability or where product owners account for the technical constraints experienced by engineers during implementation. Such interactions shape collaboration, communication, and user-centred design outcomes. Although SE research has increasingly recognised empathy as a key human aspect, there remains no validated instrument specifically designed to measure it within the unique socio-technical contexts of SE. Existing generic empathy scales, while well-established in psychology and healthcare, often rely on language, scenarios, and assumptions that are not meaningful or interpretable for software practitioners. These scales fail to account for the diverse, role-specific, and domain-bound expressions of empathy in SE, such as understanding a non-technical user's frustrations or another practitioner's technical constraints, which differ substantially from empathy in clinical or everyday contexts. To address this gap, we developed and validated two domain-specific empathy scales: EmpathiSEr-P, assessing empathy among practitioners, and EmpathiSEr-U, capturing practitioner empathy towards users. Grounded in a practitioner-informed conceptual framework, the scales encompass three dimensions of empathy: cognitive empathy, affective empathy, and empathic responses. We followed a rigorous, multi-phase methodology, including expert evaluation, cognitive interviews, and two practitioner surveys. The resulting instruments represent the first psychometrically validated empathy scales tailored to SE, offering researchers and practitioners a tool for assessing empathy and designing empathy-enhancing interventions in software teams and user interactions.

cs.SE

新论文 · v1arXiv:2611.21694

TIBECG: Temporal Information Bottleneck for Lead-Specific Demographic Privacy in ECG Waveform Release

Tsai-Min Chen, Jun Seita

原文摘要

Transmitting an electrocardiogram (ECG) waveform can expose demographic information beyond an intended diagnostic task. We evaluate TIBECG, an architectural temporal information bottleneck, on individual leads extracted from hospital twelve-lead recordings. A 2,302,337-parameter processor varies its latent length while retaining the same nine-score consistency objective, without age or sex labels in protector fitting. Across 6,386 recordings, ten folds, 12 leads and 21 lengths, retrospective selections retained at least 95% of each lead's clean macro-F1. Their weaker age/sex interval gain exceeded the best eligible conventional transformation on all leads, by a median 22.08 percentage points. The implemented sampling-rate comparison shows that direct downsampling usually suppresses demographic inference more strongly but at a larger loss of diagnostic utility, whereas TIBECG favors utility preservation. Selection stability also varies by lead. Lower latent Kullback-Leibler (KL) divergence does not reliably rank residual sex leakage at a fixed length. These findings support lead-specific evaluation of waveform processing, not validated consumer deployment or expert-readable reconstruction. When the waveform-input diagnostic classifier is hosted as part of a proprietary cloud service rather than deployed on the device, waveform transmission is required to obtain its scores.

cs.LG · cs.AI · cs.CR · eess.SP

版本更新 · v3arXiv:2607.04412

LLM-as-a-Tutor: Policy-Aware Prompt Adaptation for Non-Verifiable RL

Yujin Kim, Namgyu Ho, Sangmin Hwang, Joonkee Kim, Yongjin Yang, Sangmin Bae, Seungone Kim, Jaehun Jung 等 10 位作者

原文摘要

Reinforcement learning (RL) for non-verifiable instruction following increasingly relies on LLM judges with prompt-specific rubrics as reward signals. While recent methods adapt these rubrics to the evolving policy during training, the training prompts themselves remain static, drawn from fixed corpora. This static approach often results in a critical misalignment between prompt difficulty and policy capability, leaving the judge unable to recover a discriminative reward signal when prompts fail to elicit quality variance among rollouts. To address this misalignment, we introduce LLM-as-a-Tutor, a framework that extends the LLM's role from judge to tutor: a single model serves as an examiner that pairwise-compares policy rollouts to detect non-challenging prompts, and as a generator that appends atomic constraints to them. This append-only design preserves the original prompt while adding constraints to increase difficulty in step with the policy's capability, producing a self-calibrating training signal without external difficulty schedules. On three complex instruction-following benchmarks, our method outperforms both policy-unaware baselines and prior policy-adaptive methods that adapt rubrics or rewrite prompts, on average and on four of five metrics, suggesting prompt adaptation as an underexplored axis of policy-awareness in non-verifiable RL for constraint-heavy instruction following.

cs.AI

新论文 · v1arXiv:2611.20906

Can Image Generators Hold a Scene Together When Space Is Broken Apart?

Jhen-Ke Lin, Chung-Chun Wang

原文摘要

Image generators are good at drawing scenes in which things that belong together appear next to each other. In almost every image they generate, and in almost every benchmark used to evaluate these images, "belonging to the same whole" and "adjacent to each other" go together. Therefore, we cannot know whether the model truly grasps the scene as a whole, or simply makes each part connect with its adjacent parts. We propose GJBench to disentangle the two. It requires the model to draw a scene as a scattered jigsaw puzzle: an image is cut into interlocking pieces, then moved apart and rotated. Thus, the spatial plausibility of the scene exists only in how these pieces collectively point to a whole. The model may have a reference image, a target layout, both, or neither. We separately evaluate three requirements: whether the scene is broken down into a specified number of separate pieces, whether the pieces interlock, and whether the pieces can reconstruct the reference scene. While every open and closed model we evaluated could draw these scenes well as a whole, none could keep them together as scattered puzzle pieces, and the three requirements categorized failures into different patterns. The models draw separate pieces but miscount them in different ways, and unless their outlines are provided, the pieces rarely interlock, and their content does not reconstruct the reference image. We also found that the target layout helps the puzzle pieces fit together, while the reference image improves neither their count nor their fit. These findings show that keeping a scene continuous across space that the canvas no longer shows remains an open problem for image generators.

cs.CV

版本更新 · v2arXiv:2604.17277

Analog Resonant Recurrent Neural Networks via Metacircuits

Zixin Zhou, Tianxi Jiang, Menglong Yang, Mengli Sui, Yizhuo Zhang, Zhihua Feng, Qingbo He, Shiwu Zhang

原文摘要

Physical neural networks offer a promising route to fast and energy-efficient edge intelligence by embedding computation in the intrinsic dynamics of materials and devices. However, processing temporal signals requires recurrent networks with memory, whose trained dynamics are hard to reproduce faithfully in physical systems. Here we present an analog resonant recurrent neural network (R$^2$NN) implemented as a metacircuit of coupled local resonators. A reformulated mechanical-electrical analogy based on frequency-dependent negative resistances maps a trained network one-to-one onto the metacircuit, with every trainable parameter realized as a resistance. Training tunes local resonances and global resistive couplings so that past responses persist as recurrent memory and interfere with incoming inputs, and the resulting impedance landscape routes different frequencies to distinct outputs. Exploiting this recurrent memory, the metacircuit solves a temporal-parity task whose output depends on the entire input history. Building on the learned frequency selectivity, stand-alone R$^2$NN systems classify raw tactile, acoustic and vibration signals in real time and directly drive downstream hardware. This work establishes engineered locally resonant dynamics as a physical substrate for recurrent neural computation and opens a route to compact, low-latency neural hardware for machine intelligence at the edge.

cs.LG · cs.AI · cs.ET · physics.app-ph

版本更新 · v2arXiv:2610.35031

PatentAgentBench: Benchmarking LLM Agents with Professional Patent Task Suites

Yijie Hu, Wei Li, Junjie Zhou, Zhiqi Jiang, Yi Jin, Licong Xu, Xinkai Sun, Kai Duan 等 9 位作者

原文摘要

Patent practice lies at the intersection of technical, legal, and linguistic work, making it challenging yet crucial to technological innovation. Recently, rapid advances in large language models (LLMs) and agentic systems have demonstrated the potential to complete realistic business workflows and even patent works. This raises a question: can current AI systems perform patent work reliably? Existing benchmarks cover a limited range of patent tasks and simplify evidence acquisition by providing reference materials or candidate documents. In this paper, we introduce PatentAgentBench, a benchmark comprising 510 instances across eight task suites: novelty search, freedom-to-operate (FTO) search, design FTO, invalidity search, claim charting, patent drafting, office-action response, and patent translation. Curated from real patent records and professional materials, the benchmark combines task-specific metrics with expert-reviewed annotations and rubrics. Retrieval tasks such novelty search, FTO, etc., require agents to search and compare patent collections without a supplied candidate pool. The drafting and OAR tasks assess the quality of agents' work given technical disclosure and office action notice. We evaluate seven LLMs across four agent harnesses on PatentAgentBench, finding that current systems still struggle to identify novelty, assess risk, and understand practitioner intent throughout patent workflows.

cs.AI

版本更新 · v3arXiv:2510.23587

A Survey of Data Agents: Emerging Paradigm or Overstated Hype?

Yizhang Zhu, Liangwei Wang, Chenyu Yang, Xiaotian Lin, Boyan Li, Wei Zhou, Xinyu Liu, Zhangyang Peng 等 25 位作者

原文摘要

The rapid advancement of large language models (LLMs) has spurred the emergence of data agents, autonomous systems designed to orchestrate Data + AI ecosystems for tackling complex data-related tasks. However, the term "data agent" currently suffers from terminological ambiguity and inconsistent adoption, conflating simple query responders with sophisticated autonomous architectures. This terminological ambiguity fosters mismatched user expectations, accountability challenges, and barriers to industry growth. Inspired by the SAE J3016 standard for driving automation, this survey introduces the first systematic hierarchical taxonomy for data agents, comprising six levels that delineate and trace progressive shifts in autonomy, from manual operations (L0) to a vision of generative, fully autonomous data agents (L5), thereby clarifying capability boundaries and responsibility allocation. Through this lens, we offer a structured review of existing research arranged by increasing autonomy, encompassing specialized data agents for data management, preparation, and analysis, alongside emerging efforts toward versatile, comprehensive systems with enhanced autonomy. We further analyze critical evolutionary leaps and technical gaps for advancing data agents, especially the ongoing L2-to-L3 transition, where data agents evolve from procedural execution to autonomous orchestration. Finally, we conclude with a forward-looking roadmap, envisioning proactive, generative data agents. A curated companion repository is available at https://github.com/HKUSTDial/awesome-data-agents.

cs.DB · cs.AI

新论文 · v1arXiv:2611.20900

Joint Learning of Control Lyapunov Functions and Controllers via Stable-by-Design Surrogates

Patricia Pauli, Matthieu Barreau

原文摘要

Training and verifying neural network-based control Lyapunov functions (CLFs) can be computationally demanding. To address this challenge, we propose a framework that jointly learns a CLF and a stabilizing controller using structured neural networks and a projection-based parameterization. Specifically, we employ Polyak-Lojasiewicz networks (PLNets), which parameterize nonconvex, positive-definite CLF candidates, and construct a stable-by-design surrogate of the closed-loop dynamics that satisfies the Lyapunov decrease condition by construction. The controller combines linear state feedback with a higher-order correction term resembling Sontag's universal formula and is trained to align the resulting closed-loop dynamics with the stable-by-design surrogate. Unlike classical approaches based on Sontag's formula, our framework does not require a known CLF; instead, it learns the CLF and controller simultaneously. We establish local and global stability guarantees for the learned controller and demonstrate the effectiveness of the proposed approach through numerical examples.

eess.SY · cs.LG

版本更新 · v2arXiv:2602.05827

Sparse Video Generation Propels Real-World Beyond-the-View Vision-Language Navigation

Hai Zhang, Siqi Liang, Li Chen, Yuxian Li, Yukuan Xu, Yichao Zhong, Fu Zhang, Hongyang Li

原文摘要

Why must vision-language navigation be bound to detailed and verbose language instructions? While such details ease decision-making, they fundamentally contradict the goal for navigation in the real-world. Ideally, agents should possess the autonomy to navigate in unknown environments guided solely by simple and high-level intents. Realizing this ambition introduces a formidable challenge: Beyond-the-View Navigation (BVN), where agents must locate distant, unseen targets without dense and step-by-step guidance. Existing large language model (LLM)-based methods, though adept at following dense instructions, often suffer from short-sighted behaviors due to their reliance on short-horimzon supervision. Simply extending the supervision horizon, however, destabilizes LLM training. In this work, we identify that video generation models inherently benefit from long-horizon supervision to align with language instructions, rendering them uniquely suitable for BVN tasks. Capitalizing on this insight, we propose introducing the video generation model into this field for the first time. Yet, the prohibitive latency for generating videos spanning tens of seconds makes real-world deployment impractical. To bridge this gap, we propose SparseVideoNav, achieving sub-second trajectory inference guided by a generated sparse future spanning a 20-second horizon. This yields a remarkable 27x speed-up compared to the unoptimized counterpart. Extensive real-world zero-shot experiments demonstrate that SparseVideoNav achieves 2.5x the success rate of state-of-the-art LLM baselines on BVN tasks and marks the first realization of such capability in challenging night scenes.

cs.CV · cs.RO

新论文 · v1arXiv:2611.20894

Which AI Use Cases Deserve a Pilot? A Risk and Value Framework for Funding, Scoping, and Scaling AI

Vishal Srivastava

原文摘要

A composite score can rank AI use cases while leaving the permitted operating scope and the evidence required for expansion unspecified. We treat the unit of decision as a use case at a proposed scope (authority, human mediation, limits, volume, environment, population) and define three decisions: screen the scope, fund the experiment, and authorize expansion only when residual risk is acceptable and evidence of incremental net value applies at that scope. A risk track (R1-R4) is implemented as an explicit decision table with vetoes and documented precedence, and a value track follows four evidence stages with each transition needing its own evidence. Funding uses expected value of perfect information as an upper bound and the value of sample information for the specific experiment. A quantified expected-loss check, required where policy says so, computes the containment an expansion needs and credits a control only on evidence measured at the proposed scope; in a synthetic scenario the highest-value control design fails that test while a smaller scope passes. A reproducible Python package with 279 passing tests demonstrates the framework on synthetic refund, reconciliation and knowledge-assistant use cases, including a synthetic matched-scope fixture and a transferability stress test (value evidence is simulated, not realized), and compares it with a public AWS sample. Thresholds and appetite are prototype policy requiring field validation.

cs.AI

新论文 · v1arXiv:2611.20893

Disaggregated LLM Inference in Space: A Communication Feasibility Study

Arno Troch, Roger Sanchez-Vital, Seyyidahmed Lahmer, Bartek Kozicki, Jeroen Famaey, Michael Peeters

原文摘要

Orbital data centers could enable large-scale AI inference, but efficiently serving a frontier language model to many users may require several compute satellites to cooperate over optical links. While disaggregated inference, which splits large language model (LLM) serving across machines, is well established in terrestrial data centers, its viability under the unique constraints of orbital deployments remains an open question. This article investigates how link capacity and satellite separation affect two types of disaggregated LLM serving, in both compact satellite formations and distributed constellations. Prefill-decode disaggregation (PDD) migrates a large KV cache once per request, while attention-FFN disaggregation (AFD) exchanges small activations at every mixture-of-experts layer for every generated token. The two respond to distance in different ways. PDD is limited mainly by link capacity and, with fast enough links, tolerates separations of up to a thousand kilometers. AFD is limited mainly by propagation delay, which accumulates layer by layer and forces attention and expert instances within about a kilometer of each other. Based on these results, we propose a hierarchical orbital inference architecture and provide network design guidelines for future orbital data centers, including the role of ground access latency.

cs.NI

新论文 · v1arXiv:2611.20891

From Recursion to Convolution: Parallel Self-Improvement for Multi-Agent Systems

Zheng Wang, Yuang Liu, Sachith Abeywickrama, Yangkai Ding, Tao Yu, Xinyu Zhang, Hongbo Wang, Qi Ye 等 10 位作者

原文摘要

Multi-agent self-improvement enables agents to improve through interaction and feedback. However, sequential dependencies across agents and refinement rounds can propagate errors, increase computational costs, and limit parallelism. Moreover, refining individual features or agents in isolation can overlook their dependencies and restrict collective coordination. To address these challenges, we propose ConvRSI, a framework for parallelizing multi-agent self-improvement through convolution. ConvRSI couples intra-agent convolution, which jointly refines related agent features, with inter-agent graph convolution, which aggregates messages according to the collaboration topology. Both stages operate in latent space, enabling parallel updates without repeatedly decoding and re-encoding intermediate messages. Experiments on MATH-500 show that ConvRSI achieves 5.04% performance improvement while reducing 17.27% inference cost, and 45.11% latency compared with RecursiveMAS. It also supports diverse collaboration patterns by representing their topologies as adjacency matrices. These results demonstrate the potential of convolutional co-evolution for efficient multi-agent self-improvement. Our code and datasets are provided in https://github.com/LYA2333/ConvRSI.

cs.AI · cs.CL · cs.LG

版本更新 · v2arXiv:2610.23909

When the Editing Intent Is Split: A Cross-Modal Jailbreak Attack for Large Image Editing Models

Xianmin Ye, Zhiyuan Fan, Tianju Liu, Yuhang Wang, Zhenxing Niu, Haichang Gao

原文摘要

Large image editing models increasingly infer editing intent from both textual and visual inputs, creating an underexplored safety risk when harmful intent is distributed across modalities. In this work, we formalize this split-intent vulnerability and propose the Split-Intent Jailbreak Attack (SIJA), which asymmetrically distributes an unsafe editing intent by specifying the editing referent in text while concealing the safety-critical action and specification through graphical visual semantics in an auxiliary cue. To systematically study this threat, we introduce the Split-Intent Image Editing Safety Benchmark (SI2ESBench) for evaluating image editing models under split-intent inputs. Extensive experiments on representative commercial and open-source models demonstrate that SIJA effectively compromises state-of-the-art image editing systems, achieving ASRs of 68.28% on GPT Image 2 and 91.38% on Qwen-Image-3.0-Pro. Our results expose a mismatch between multimodal editing and safety assessment: the distributed intent can be recovered to execute the requested transformation, yet its harmfulness often remains undetected by existing safeguards. To mitigate this vulnerability, we propose editing-intent reconstruction, a defense principle that explicitly evaluates the complete transformation implied by the joint multimodal context and substantially improves rejection of split-intent attacks. Our findings expose a cross-modal safety gap in modern image editing systems and highlight the need to assess the jointly inferred transformation rather than individual inputs in isolation. Project page: https://github.com/yexianmin/SIJA.

cs.CV · cs.AI

版本更新 · v4arXiv:2610.29271

ARCANA: A Benchmark for Abstraction and Analogical Reasoning

Ruqi Chen, Zilei Wang, Wei Wang

原文摘要

Analogical reasoning is widely considered a fundamental component of human intelligence. Few benchmarks are designed to explicitly test far analogical reasoning - reasoning between domains that are superficially dissimilar yet share deep relational or structural similarities. In this work, we introduce an analogical reasoning benchmark called Abstraction and Reasoning Corpus of ANAlogy (ARCANA). The tasks of ARCANA require performing far analogical mappings from real-world objects or physical phenomena to abstract grid-based transformations, different from the well-known ARC-AGI benchmark which primarily emphasized innate human priors. We validate ARCANA through a human study, demonstrating that humans solve these tasks reliably. In contrast, evaluations across a range of large language models show a distinct performance gap compared to humans, revealing systematic failure modes that remain challenging for current models. To address this gap, we propose ConceptSIR, a reasoning framework based on concept slippage, introspection, and reflection. Experiments demonstrate that ConceptSIR enables standard LLMs to solve several challenging ARCANA tasks that were unsolved by direct prompting in our evaluation. Overall, our results suggest that far analogical reasoning remains an open problem for contemporary LLMs. All the materials are available at sites.google.com/view/arcanacogtest. The complete codebase, including benchmark, and ConceptSIR, is publicly available at https://github.com/density-11/arcana-test. All ARCANA data and task instances can be accessed and explored at https://arcanacogtest.org/.

cs.AI

版本更新 · v3arXiv:2602.07834

What Does Machine Learning Learn from Numerical Calabi--Yau Geometry? A Controlled Study on the Fermat Quintic

D Yang Eng

原文摘要

We study polynomial regression of numerical metric errors on the Fermat quintic. A known hypersurface volume-ratio identity gives the exact reference $C-\log p_4$ relative to the induced Fubini--Study metric, where $p_4=\sum_i|z_i|^8$ on unit-normalized homogeneous coordinates. We optimize algebraic metrics at degrees $6,8,10,12$ while preserving finite Fermat symmetry, then fit their remaining logarithmic volume-form error. Regression splits and bootstrap draws keep the five roots of each sampled line together, with all compared teachers evaluated on common clouds. Warm-start replication correlations exceed $0.9997$ at $k=8,10,12$. Across degrees, source-derived modes fall short of target-specific fits, including in a post-hoc quartic check, but outperform random spans of matched rank in the tested libraries. Two independent random-invariant $k=10$ starts have residual correlations $0.768$--$0.803$, while their fitted cubic modes correlate at $0.979$--$0.983$ and retain useful calibrated cross-start prediction. Their fitted modes agree poorly with those of the principal warm-start teacher. Adding the phase-sensitive invariant $J_7$ improves prediction beyond the cubic modulus-only library for all seven warm-start teachers. The gain depends on the library: at $k=12$, a quartic modulus-only fit explains about 97--98\% of residual variance and the conditional interval for the added $J_7$ gain includes zero for both replicas under both measures. The gains therefore establish neither phase independence nor irreducible phase necessity; fixed-baseline augmentation preserves this distinction. The fitted components depend on the numerical protocol; global optimality and a universal finite-degree correction remain unestablished.

cs.LG · math.DG

新论文 · v1arXiv:2611.20885

ProSE: Probing Reconstruction Error for Semantic-guided Expert Skipping in MoE-MLLMs

Jiameng Li, Minye Wu, Matthew B. Blaschko

原文摘要

Mixture-of-Experts Multimodal Large Language Models (MoE-MLLMs) offer scalable inference through sparse expert activation, yet substantial computational redundancy remains among the activated experts. Existing expert skipping methods typically calibrate thresholds based on target skipping ratios or final-output divergence, overlooking the accumulation of skipping-induced error across layers and the varying semantic importance of individual tokens. To address that, we propose ProSE, a training-free, semantic-guided expert skipping framework that probes layer-wise reconstruction error to promote representation consistency compared with default activation. Specifically, we construct query-aware semantic masks to identify salient visual tokens and progressively calibrate modality-specific skipping thresholds for each MoE layer. By minimizing masked reconstruction error under a given budget, ProSE maintains semantic consistency after skipping and allows adaptive expert allocation under limited activation budgets. Extensive experiments on 9 benchmarks demonstrate consistent improvements on average over existing expert skipping methods, particularly under aggressive skipping settings. ProSE maintains over 96\% performance on KimiVL-A3B-Inst with 83\% expert skipping and accelerates prefilling by $1.7\times$.

cs.CV

新论文 · v1arXiv:2611.20879

Multi-Rate Flow Matching for Point Cloud Upsampling

Jingkai Shang, Edgar Torres, Duy Nguyen

原文摘要

Flow-matching point-cloud upsamplers refine sparse inputs into dense point clouds by integrating a learned velocity field with a fixed temporal discretization. However, local reconstruction requirements vary substantially across a shape, while a shared solver budget applies the same temporal resolution everywhere. We propose Multi-Rate Flow Matching (MRFM), a patch-adaptive framework that selects an integration budget for each local patch at inference time while keeping the learned velocity field frozen. MRFM relaxes the discrete budget into a continuous variable, optimizes it online through differentiable soft rollouts using an input-derived geometric fidelity objective, and projects the result back to a discrete budget for the final solver rollout. Experiments on PU1K and PUGAN show strong reconstruction quality, while matched-budget permutation and multi-factor analyses demonstrate that the benefit depends on patch-specific allocation rather than on the step distribution alone. MRFM is accuracy-oriented and intentionally uses additional test-time optimization to obtain more suitable local temporal resolutions.

cs.CV

新论文 · v1arXiv:2611.20877

Markov-Equivalent Conditional Independence Graphs on Path Spaces

Simon Kempf, Niki Kilbertus, Georg Manten

原文摘要

Causal discovery for continuous-time stochastic processes with latent confounding typically faces a trade-off between (i) approaches that are highly informative about the causal structure but are difficult to test statistically, and (ii) approaches that admit practical tests from data but yield more coarse-grained conclusions, potentially not leveraging the direction of time. We introduce W-separation, an asymmetric graphical criterion for directed mixed graphs defined as a quaternary relation that allows distinct conditioning on past and future components. This added flexibility yields a highly informative graphical independence model while preserving testability. W-separation satisfies a global Markov property for a broad class of stochastic differential equations with latent confounding, linking it to conditional independence relations among coordinate processes that are testable from data. A central result of this work is the full characterization of W-Markov equivalence classes: we show that every equivalence class admits a unique greatest element that serves as a canonical representative and a principled target for learning algorithms. In addition, we provide necessary and sufficient conditions under which directed or bidirected edges can be added without altering the induced independence model.

cs.LG

新论文 · v1arXiv:2611.20873

Re-Reading Is Not Free: Token Traces and Cache-Aware Energy Estimation for LLM Agents in Software Engineering

Tina Vartziotis, Ketevan Peranidze, Rodopi Kosteli, Ezgi Sarikayak, George Dasoulas, Michael Keckeisen, Sotirios Kotsopoulos, Francesca Dominici

原文摘要

As large language models (LLMs) evolve from conversational assistants into autonomous agents embedded in software development workflows, estimating their environmental footprint requires moving beyond conventional token-based accounting. Existing energy estimators focus on isolated request-response interactions, overlooking the repeated processing of accumulated context, cache reuse, and work delegated to subagents. We analyze nearly one billion tokens collected across 42 sessions and 752 task episodes over ten weeks of real-world AI-assisted software development. Cached-context reads make up 94.8% of all API-reported tokens, while uncached input accounts for only 0.017%; for every generated token, approximately 260 cached tokens are read. Context grows 5.3-fold over a session compared with 1.4-fold within individual tasks. To quantify the energy implications, we extend an existing analytical inference-energy model to account for cached-context processing, yielding a cache-aware hybrid estimator. It estimates approximately 250 kWh for the workload, compared with approximately 10 kWh from the output-based estimator EcoLogits, a 25-fold difference largely driven by cached-context reads. Neither estimator has been validated on long cached contexts, but no modeling assumption we varied closes the gap. Starting each task from a fresh context would lower our estimate by about 60%. We also find that parent-session logs report only a subagent's final internal call, undercounting delegated tokens by a factor of ten; for this workload the correction changes the energy estimate by about 5%, but token-volume studies are fully exposed to it. Our findings expose a substantial gap between how current energy estimators and agentic workloads treat context, highlighting the need for workflow-level energy accounting that captures context accumulation, cache reuse, and delegated execution.

cs.SE · cs.AI

新论文 · v1arXiv:2611.20872

MVSTF: Multi-View Spatio-Temporal Fusion for LiDAR-Based Moving Object Segmentation

Jiaoju Zhou, Chao Ye, Xuebo Yang, Weiyang Lin

原文摘要

Autonomous mobile robots confront significant chal lenges in recognizing the motion properties of objects in dy namic environments. However, existing methods are hindered by limitations such as boundary ambiguity, occlusion, viewpoint constraints, and training complexity. In this paper, we propose an innovative end-to-end Multi-View Spatio-Temporal Fusion network, named MVSTF, which utilizes sequential point clouds to generate multi-view images and feature differences, and designs a multi-branch network to extract spatio-temporal and dimen sional features in parallel, and realizes high-precision moving object segmentation. First, sequential point clouds are utilized to generate multi-view images and feature differences. Then, an end-to-end multi-branch network including Dual Down-sampling Blocks (DDB), Spatial Attention Blocks (SAB), Multi-Kernel Convolutional Neural Network blocks (MKCNN), Dual Feature Fusion Blocks (DFFB), and the Point Fusion (PF) module, is designed to extract motion features from feature differences and geometric and semantic features from multi-view images. Finally, an auxiliary semantic head is introduced to further improve the algorithm's robustness and the model's interpretability. This work is validated on the SemanticKITTI and KITTI-Road bench marks, achieving state-of-the-art results on the SemanticKITTI MOS validation set and the KITTI-Road validation set, and competitive results on the hidden SemanticKITTI-MOS test set. The code is available at https://github.com/ZhouJJ111/MVSTF.

cs.CV · cs.RO

版本更新 · v2arXiv:2610.24135

Inside the Loop: A Mechanistic Study of Looped Transformers on Permutation Traversal

Magnus Sesodia, Juil Sock, Philip Torr, Christian Schroeder de Witt

原文摘要

Solving algorithmic tasks with many sequential steps typically requires scaling up transformer depth or chain-of-thought length, incurring unfavorable parameter and inference-compute costs, respectively. Looped transformers, which reuse one block of layers across many iterations, are poised to address this, since they scale effective depth without adding parameters or tokens. Yet, it remains unclear what each iteration is actually computing, or whether the loop merely simulates a deeper untied stack. Here, we study directed permutation traversal, in which a model follows H hops from a starting node and reports the node it reaches, so every intermediate hop has a known ground truth. We find that a minimum effective depth is necessary to solve the task, beyond which the degree of weight tying becomes the determining factor. For instance, a looped transformer with a four-layer block iterated eight times solves the task in every seed, while an identically trained untied transformer of the same effective depth never does, succeeding only with a lower learning rate or supervision of its intermediate layers. Probing and intervening on the residual stream, we find that the looped model advances through the intermediate nodes gradually, forming a staircase that adapts to its iteration budget, whereas the untied model, at that lower rate, exposes them only in its final layers. Taken together, our results recast looped transformers as iterative program learners, which allocate their iterations to successive steps of the algorithm and solve tasks that identically trained untied models of the same effective depth do not. Code is available at https://github.com/magnus-sesodia/insidetheloop.

cs.LG · cs.CL

新论文 · v1arXiv:2611.20868

It Depends on the Task: Rethinking Rhetorical Segmentation of Dutch Judicial Decisions

Samyak Sheth, Gerasimos Spanakis, Gijs van Dijck

原文摘要

Classifying the material facts, submissions, judicial reasoning, and the court's decision (rhetorical roles) in case law is a prerequisite for high-quality performance across virtually all NLP-in-law tasks, including legal information retrieval, summarisation, and LLM-based analysis. These rhetorical roles often intertwine and shift sentence by sentence, making reliable classification difficult. This paper introduces the first sentence-level annotated dataset of Dutch judicial decisions, labelled across four legally meaningful categories: material facts, submissions, judicial reasoning, and the court's decision, alongside non-annotated (None) content, and compares several classification strategies, including a staged pipeline that separates None detection from the rhetorical roles, and a Bayesian update that leverages information about the section-header a sentence belongs to. Our results show that no single strategy dominates on both precision and recall: header priors help most when the None sentences are excluded and offer little benefit for None detection itself. These trade-offs matter in practice. For instance, extractive summarisation depends more on recall so it does not miss relevant facts, whereas outcome prediction depends more on precision regarding decision and facts sentences so that the predicted ruling is not contaminated by noise. We use these trade-offs and map the approaches tested in this study onto a set of representative legal NLP use cases.

cs.CL