category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.06945v1 Announce Type: new Abstract: Mechanistic interpretability often relies on the Linear Representation Hypothesis (LRH), which assumes that high-level concepts are encoded as linear directions in activation space. Yet a natural visual concept does not necessarily require a linear visual transition: between sunny and stormy lies an intermediate weather state such as a sky with a few
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07032v1 Announce Type: new Abstract: Large-scale pretrained transformer models have achieved state-of-the-art performance across diverse machine translation tasks, including multilingual settings. Knowledge distillation has emerged as a sustainable approach for model compression, transferring knowledge from large teacher models to smaller, more efficient student models. Similarly, quant
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.06932v1 Announce Type: new Abstract: Cache-based test-time adaptation (TTA) for vision-language models is often hindered by background bias in global representations and unreliable entropy-based cache admission under representation variations. To address these limitations, we propose RADC, which enhances prototype learning through reliable dual caching. RADC introduces a Semantic Foregr
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.06938v1 Announce Type: new Abstract: Medical image segmentation remains fragmented along two axes: segmentation paradigms and data dimensionality. Existing methods are typically developed separately for semantic, in-context, and interactive segmentation, and are further specialized to either native 2D images or 3D volumetric data. In clinical practice, however, segmentation workflows ta
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.03948v1 Announce Type: new Abstract: Autonomous driving decision systems must balance safety, efficiency, and social norms in complex traffic interactions. Philosophical and ethical considerations have received limited attention in existing autonomous driving decision-making approaches based on numerical optimization, sequence prediction, and large language models (LLMs). We propose Chi
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.03938v1 Announce Type: new Abstract: Agentic multimodal large language models (MLLMs) have recently pushed the frontier of visual reasoning by calling tools such as zooming and tagging. Despite the recent strong success of agentic MLLMs, this work uncovers a critical safety failure in the tool-use paradigm: agentic tool-using MLLMs become less capable of refusing harmful requests. Our e
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.03934v1 Announce Type: new Abstract: Existing analog integrated circuit design benchmarks make two questions hard to answer: whether a model has learned transferable circuit skills rather than recalled familiar examples, and whether its output works under defined process and test conditions. We introduce a benchmark built from human-designed, open-source circuits associated with Tiny Ta
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07109v1 Announce Type: new Abstract: When an LLM judge scores an output, its score distribution retains uncertainty and disagreement information that is lost after scalar compression. We introduce JudgeMoE, a lightweight aggregator that assigns example-specific weights to cached judge score distributions and fuses them before computing a final score. A protocol study shows that score-ra
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.07059v1 Announce Type: new Abstract: Yield forecasts help planners and farmers decide on inputs, storage and imports, but small agricultural tables can make reported accuracy fail on a new season. We built a crop-yield prototype for Punjab, Pakistan that combines Random Forest, XGBoost, support vector regression and a Ridge-stacked ensemble with a MobileNetV2 leaf-health classifier, and
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.03894v1 Announce Type: new Abstract: A deployed LLM agent emits tool calls, queries, and code that can be silently wrong -- by the time the error surfaces, the action has run. Frontier chat APIs hide the model's token probabilities; the agent's stated confidence barely beats chance on the mistakes that matter; and resampling does not help, since frontier models are highly repetitive, re
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04053v1 Announce Type: new Abstract: Autonomous agents built on Large Language Models (LLMs) need standardized protocols to interoperate across systems. Several now exist (A2A, MCP, ACP, ANP, NLIP), but the Natural Language Interaction Protocol (NLIP) has not appeared in any controlled performance study, and no work has measured where an agent protocol's latency is spent. We compare NLI
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04021v1 Announce Type: new Abstract: Generative artificial intelligence has become increasingly incorporated into digital media and more generally into production workflows with which the public frequently interacts. Current provenance standards and disclosure methods frequently rely on binary categorizations, differentiating only between entirely human-authored and AI-generated content
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.06896v1 Announce Type: new Abstract: Frontier multimodal large language models (MLLMs) are increasingly positioned as general purpose visual reasoners as part of the quest for artificial general intelligence. A key test of this generality is whether they can perform novel visual judgments that humans can make reliably from visual evidence and task instructions, without task-specific par
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04056v1 Announce Type: new Abstract: Synthesizing realistic graphs at scale is vital when the graphs of interest are large and real-world samples are limited or access-sensitive. Diffusion-based generators have recently driven much of the progress, offering high modeling capacity, but most such methods have quadratic computational complexity and are hence restricted to small-scale netwo
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.03888v1 Announce Type: new Abstract: Reconstructing global sea surface pH from sparse observations is critical for monitoring ocean acidification and understanding marine carbon cycling. Traditional assimilation and inverse models are physically grounded but costly for large-scale reconstruction. Recent black-box and physics-guided AI models improve efficiency, but are mainly designed f
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.07031v1 Announce Type: new Abstract: Recent video world models have witnessed the paradigm shift from single-agent to multi-agent involvements, which can reveal more complicated dynamics and cross-agent interaction in the real world. However, existing approaches commonly adopt implicit inter-agent communications via cross attention, which lack explicit geometry constraints and unified 3
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.07025v1 Announce Type: new Abstract: Multi-person 3D pose estimation with WiFi channel state information (CSI) is challenging because reflections from different people overlap without directly identifying individual joints. Existing masked embedding objectives capture wireless relationships without explicit pose supervision, while structured decoders can retain coordinate errors. We pro
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07643v1 Announce Type: new Abstract: Most KV-cache eviction methods ask, in effect, which memory appeared important while reading the prompt? We instead ask, which memory will matter while answering? Since decoding queries are unavailable at eviction time, prior future-aware methods rely on pseudo-responses or synthetic future-query estimates. We cast fixed-budget future-aware eviction
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.07008v1 Announce Type: new Abstract: Few-shot class-incremental learning (FSCIL) aims to learn novel classes from limited annotations while preserving prior knowledge. Existing methods typically assume a sufficiently large base session, but this assumption fails when both base and incremental data are scarce, leading to weak initial representations, semantic drift, and unstable boundari
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04040v1 Announce Type: new Abstract: Financial LLM agents are often evaluated by comparing their end-to-end returns with those of a baseline and testing the paired difference against zero. This measures whether deploying the agent changes realized performance, but it does not isolate event-selection skill. An agent that frequently changes positions from flat to long can earn a positive
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06980v1 Announce Type: new Abstract: Sensing technologies have advanced rapidly across industries ranging from energy to automotive manufacturing. These systems generate high-dimensional (HD) data characterized by complex nonlinear patterns and strong temporal dependencies. Traditional statistical monitoring methods are often limited in their ability to capture such nonlinear structure.
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06918v1 Announce Type: new Abstract: We study federated reinforcement learning in which multiple agents interact with a common Markov decision process and communicate through a central server to collaboratively learn the optimal state-action value function. Our goal is to understand whether the sample-efficiency benefits of collaboration can be retained when a fraction of the agents beh
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07659v1 Announce Type: new Abstract: Speculative decoding accelerates autoregressive generation in large language models. In each drafting stage, a lightweight draft model proposes tokens that the target model subsequently verifies. With increasingly capable draft models, we find that the target model frequently accepts all tokens produced in a drafting stage. A verification nevertheles
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.06903v1 Announce Type: new Abstract: Activation steering manipulates large language model behavior by intervening on internal activations, but the mechanistic basis of these interventions remains poorly understood. We decompose refusal steering into component-level interventions across four open-weight models, identifying the sparse subsets of attention and MLP components whose steering
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.06902v1 Announce Type: new Abstract: Retrieval-augmented generation grounds language models in external context, but for long documents flat top-$k$ retrieval can cluster on a single region and miss complementary evidence. RAPTOR-style summary trees address this by recursively clustering chunks and using a language model to summarize each cluster at indexing time, then ranking summary n
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06970v1 Announce Type: new Abstract: Egocentric bimanual hand pose estimation is important for virtual interaction, wearable control, and rehabilitation, but visual observations are often degraded by self-occlusion, hand-hand contact, and object manipulation. We propose EVFormer, a multimodal framework that combines the current RGB frame with the preceding 200 ms of bilateral wrist surf
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06989v1 Announce Type: new Abstract: Pavement agencies must translate a spatially distributed distress inventory into a bounded, actionable repair-lot plan: accident-critical defects (potholes) must always be addressed, lower-risk defects (cracks) should be included only when their benefit justifies the repair cost, and historical patch locations signal re-degradation risk without thems
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06988v1 Announce Type: new Abstract: Knowledge tracing (KT) models are predominantly evaluated using aggregate metrics such as area under the curve (AUC) and accuracy. However, these global scores obscure where the remaining errors originate and fail to indicate whether a benchmark is approaching saturation. While estimating a global theoretical performance limit is challenging in reali
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07700v1 Announce Type: new Abstract: We study the robustness of keystroke dynamics for detecting large language model (LLM)-assisted writing. We introduce a Vietnamese keystroke dataset capturing realistic writing modes, including bona fide composition, transcription, and paraphrasing. We also define a behaviorally grounded threat model in which users deliberately alter typing patterns.
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06974v1 Announce Type: new Abstract: Knowledge graph embedding (KGE) methods represent entities and predicates in continuous vector spaces to infer missing knowledge. Despite strong benchmark performance, their predictions often lack principled reliability guarantees, limiting their use in high-stakes applications. Moreover, uncertainty arises throughout the KGE pipeline, from incomplet
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04112v1 Announce Type: new Abstract: Sparse Autoencoders (SAEs) decompose model activations into sparse combinations of interpretable dictionary atoms. Although SAEs are grounded in the Linear Representation Hypothesis (LRH), their objective smuggles in an additional prior: concepts across patches are treated as independent, an assumption clearly violated by natural images and by the ac
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07716v1 Announce Type: new Abstract: Prefill-only decision models inspired by the Jev model score every candidate in a menu during a single forward pass and never decode, which makes one call one to two orders of magnitude cheaper than a same-scale generative language model. We show that this read-out structure comes with a testable property. When an intervention changes only the candid
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06952v1 Announce Type: new Abstract: Neural PDE solvers can achieve low prediction errors, but do they reproduce the dynamics of the systems they model? Prediction scores alone offer an incomplete answer: they measure agreement with reference solutions but provide limited insight into how errors accumulate, nearby states diverge, or extreme events arise. We propose an evaluation framewo
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04116v1 Announce Type: new Abstract: The prevailing approach to computer-use agents couples a model with a domain-specific harness: a browser or desktop environment equipped with human engineered tools that are fixed before task execution. As models' coding capabilities improve, the GUI native and static harness prevents them from direct programmatic operation on system state, as well a
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04188v2 Announce Type: new Abstract: Recent studies show that artificial intelligence (AI) with language and vision capabilities still experiences limitations in spatial reasoning. In this paper, we have studied the spatial capabilities of advanced generative AI to understand the rotations of objects in 3D space, utilizing AI's image processing and language processing features. We train
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.07127v1 Announce Type: new Abstract: Recent advances in multimodal foundation models yield strong performance on static perception and reasoning benchmarks, yet such evaluations largely overlook a central aspect of intelligence: acting competently in dynamic environments over extended time horizons. We introduce PlaySuite, a large-scale benchmark for evaluating interactive visual intell
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.07060v1 Announce Type: new Abstract: Despite substantial progress in short-to-medium-range weather forecasting, predicting high-impact events such as flash droughts remains a key challenge for both early warning operations and physically-based subseasonal-to-seasonal (S2S) prediction systems. Here we demonstrate that, for S2S soil-moisture forecasting over Europe, forecast skill depends
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.06950v1 Announce Type: new Abstract: Injecting skills into a frozen language model currently costs a million parameters and a reinforcement-learning pipeline. We introduce \method{}, a System-1 decision operator trained by behavior cloning that lowers this cost by roughly two orders of magnitude. The default operator uses 330K parameters to match a 1.33M-parameter operator trained with
10月7日 04:00
category.学术arXiv cs.LG (机器学习)
arXiv:2610.07062v1 Announce Type: new Abstract: Large language models are increasingly used to simulate how individuals respond to new situations, yet the behavioral reasoning behind these responses is either inherited from pretraining or learned from individual-level annotations, which offer limited behavioral diversity and little supervision of the reasoning itself. We propose to learn behaviora
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04168v1 Announce Type: new Abstract: Agentic large language model (LLM) systems are commonly implemented as an LLM in a loop with Planning, Memory, Tools, and Control Flow. This application-focused view connects agentic LLM research with deployable systems and leaves open how such systems should be evaluated beyond end-to-end task success. Building on this view, we define agentic cognit
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.06956v1 Announce Type: new Abstract: Large speech language models have demonstrated strong capabilities in unified cross-modal understanding and generation, yet paralinguistic cues, especially emotion, remain difficult to preserve. Existing systems typically rely on entangled acoustic representations, which allow the underlying language model to depend excessively on recovered lexical c
10月7日 04:00
category.学术arXiv cs.CV (计算机视觉)
arXiv:2610.07269v1 Announce Type: new Abstract: Cross-view geo-localization is commonly solved as an image retrieval problem, matching a ground-level image against a database of satellite tiles through a jointly trained embedding. Such models are accurate, but they need large paired supervision and cannot show what evidence supports a match. In this paper, we study a different question: how much o
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04178v1 Announce Type: new Abstract: Agent harnesses are systems that coordinate model calls, tool use, and task execution to help large language models complete complex tasks. To meet task requirements and address failures, these systems are often iteratively refined by amending and patching their instructions, tools, and workflows, continuously increasing harness complexity. It is the
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04183v1 Announce Type: new Abstract: Language models exhibit remarkable robustness, continuing to produce coherent text even when their activations are perturbed by interventions like linear steering. We hypothesize that this robustness is a result of passive dynamics, i.e., constraining mechanisms in the forward pass that funnel activations toward "good" regions that produce coherent o
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.06897v1 Announce Type: new Abstract: Localizing latent structures in the activation space of language models (LMs) is central to understanding and controlling their behavior. Yet, localized structures can differ substantially in their causal influence, raising the question of what makes a structure actionable. We tackle this question by casting causal influence as a product of three fac
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04184v1 Announce Type: new Abstract: Recursive self-improvement (RSI) relies on evaluation feedback to assess progress and guide further research, yet repeatedly running complex benchmarks is costly and slows iteration. Human experts reduce this cost by selecting benchmark subsets or designing compact suites. We ask whether AI agents can automate this design process and introduce EvalRe
10月7日 04:00
category.学术arXiv cs.AI
arXiv:2610.04129v1 Announce Type: new Abstract: We introduce InvestigationWorlds, an agentic environment for legal investigation. We build on an underused artifact of U.S. civil litigation: the summary judgment motion. This motion relies upon a record composed of real evidence exhibits, and results in a court-adopted hypothesis that is treated as ground truth for the purposes of deciding the motio
10月7日 04:00
category.学术arXiv cs.CL (自然语言处理)
arXiv:2610.07426v1 Announce Type: new Abstract: Accent classifiers are typically trained with a fixed label inventory and cannot accommodate new accent categories as new data becomes available. Moreover, accented speech corpora often exhibit substantial class imbalance and/or domain shift due to differences in recording conditions across corpora. We present AccentCL, a class-incremental learning f
10月7日 04:00