独立 AI 资讯库
从自动抓取数据库实时读取，适合快速浏览、检索和后续扩展成完整资讯站。
arXiv:2609.04442v1 Announce Type: new Abstract: Large language models deployed in high-stakes settings frequently generate plausible but ungrounded claims. Standard retrieval-augmented generation (RAG) pipelines offer limited remedy, since they retrieve isolated passages without tracking cross-docu
arXiv:2609.04384v1 Announce Type: new Abstract: Chinese online comments often convey social meaning through indirect and playful language that is hard to interpret without context. Existing evaluations largely organize items around predefined phenomena or controlled pragmatic categories, leaving op
arXiv:2609.04336v1 Announce Type: new Abstract: Medical visual question answering (Med-VQA) is often assumed to require medical fine-tuning, large models, or complex multi-agent pipelines. We revisit this assumption with \textbf{MedProb}, a lightweight probing framework that predicts multiple-choic
arXiv:2609.04290v1 Announce Type: new Abstract: Despite increasing reliance on LLMs that reason with external evidence supplied by tools, retrieval-augmented generation, other agents, and users, how LLMs integrate such evidence into decisions they have already begun to form remains largely unclear.
arXiv:2609.04651v1 Announce Type: new Abstract: This paper studies adaptive recommendation under intent drift, where feedback from each recommendation outcome can reveal whether the relational evidence used for ranking is useful, missing, or misleading. While Knowledge Graphs (KGs) provide essentia
arXiv:2609.04641v1 Announce Type: new Abstract: Natural-language-to-SQL systems have ad- vanced rapidly on academic benchmarks, yet production enterprise schemas exhibit graph- like, semi-structured, deeply nested structure that current benchmarks do not measure. We make two complementary contribut
arXiv:2609.04629v1 Announce Type: new Abstract: A runtime gate for an LLM tool agent is usually cast as a filter. In a ReAct loop a rejected proposal is followed by another at the same state, so the gate is a search operator over the proposal stream whose admission criterion shapes which trajectori
arXiv:2609.04611v1 Announce Type: new Abstract: LLM agents are rapidly becoming production software, deployed to handle customer service, adjudicate disputes, and operate internal systems. Notably, the work of building them is increasingly handed to coding agents, yet existing benchmarks say little
arXiv:2609.04579v1 Announce Type: new Abstract: Grounded language-model pipelines can be divided into three stages: selecting an object, retrieving passages for it, and using that evidence to answer. If the selected object must reach the reader, losing it breaks the handoff. Benchmark recall checks
arXiv:2609.04565v1 Announce Type: new Abstract: Large language models demonstrate increasingly strong reasoning capabilities through effective post-training. Yet, prevailing post-training methods optimize over massive numbers of tokens, implicitly assuming that effective learning must be token-inte
arXiv:2609.04564v1 Announce Type: new Abstract: Self-driving laboratories (SDLs) combine automated experimentation with adaptive decision-making to accelerate scientific discovery. Their operation nevertheless often depends on human specialists who translate scientific objectives into executable cl
arXiv:2609.04523v1 Announce Type: new Abstract: Designing and authoring high-performance custom kernels for accelerators is a complex task that requires deep hardware-level expertise. Large Language Models (LLM) can be leveraged together with real-time compiler feedback to build agentic systems for
arXiv:2609.04518v1 Announce Type: new Abstract: Agent reinforcement learning (RL) increasingly runs through full execution harnesses, and a multi-harness recipe mixes two choices: exposing the policy to several harnesses, and comparing their rewards inside one relative-advantage group. We isolate t
arXiv:2609.04504v1 Announce Type: new Abstract: Cardiac, neural, behavioral, and speech measurements from wearable and mobile devices provide partial, noise-sensitive views of physiological state. BioSync combines these measurements into the \textbf{BioSync Index (BSI)}, a continuous composite digi
arXiv:2609.04495v1 Announce Type: new Abstract: We formulate indirect prompt injection as a test-time search over a task-dependent attack surface induced by the environment, user task, and injection task. To operationalize this formulation, we introduce an agentic attacker with a dedicated search h
arXiv:2609.04476v1 Announce Type: new Abstract: Performance modeling is central to hardware design and software optimization, yet constructing these models requires structured reasoning about computation, data reuse, storage, and movement. We introduce PerfReasoning, a benchmark that evaluates LLMs
arXiv:2609.04444v1 Announce Type: new Abstract: Benchmarks for the side effects an agent causes on the way to a goal already exist, but HarvestBench is the first to put a price on avoiding the side effect and to name that side effect as a living creature. It is a farm simulation: LLM sub-agents dri
arXiv:2609.04377v1 Announce Type: new Abstract: Enterprise AI deployments fail not from model inadequacy, but because organizations lack a structured substrate encoding how they decide, negotiate, and execute. Generic LLMs carry no firm-specific ontological priors; RAG remains brittle, with no path
arXiv:2609.04373v1 Announce Type: new Abstract: Large language models (LLMs) are being deployed at scale in consequential real-world systems, from financial markets to content moderation to hiring. We show that improving individual model capability can degrade rather than improve system-level outco
arXiv:2609.04343v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for consequential decisions, making their explanations an important tool for auditing model behavior. Unfortunately, these explanations can be unfaithful, failing to reflect the actual reasoning under
arXiv:2609.04304v1 Announce Type: new Abstract: We present Iris-mini and Iris-pro, two search agents trained at the 35B-A3B and 397B-A17B scales, together with the data pipeline and training recipe behind them. Tasks are reverse-constructed from the hyperlink structure of a web corpus: we author mu
arXiv:2609.04298v1 Announce Type: new Abstract: Evaluating agents on the growing number of agentic benchmarks is challenging because they often require complex environments and agent integrations. We introduce Harbor Adapters, a unified evaluation infrastructure for agentic benchmarks. Our work mak
arXiv:2609.04286v1 Announce Type: new Abstract: Artificial intelligence in recruitment has shifted the object being automated from profile pairs and ranked lists to multi-stage workflows that retrieve evidence, compare candidates, and support or execute actions. This systematized narrative review t
本文聚焦 Agent 隐私与安全领域尚未解决及不断涌现的问题，并从情境化视角展开讨论。所提供原文内容为致谢部分，介绍了报告与相关研讨会的主要组织者、撰写者和章节负责人。该工作汇聚了超过 50 位研究人员、学术合作伙伴及行业领袖，并获得多位来自学界和业界人士的支持。此类跨学科协作反映出，Agent 系统的隐私与安全挑战需要技术、治理与社会情境等多方面共同研究。
研究指出，多模态大语言模型即使能够正确报告空间事实，也不一定能在后续推理中实际利用这些事实。论文提出包含 23,196 个输入的 \textsc{SpaceConflict} 基准，用于评估空间状态的构建与使用，覆盖局部事实绑定、关系组合、跨观察一致性及变换下的状态判断四个层级。实验发现，模型能够从视觉证据中恢复初始状态，却常在需要利用该状态完成变换推理时失败；例如 Qwen3.5-9B 在 100 个初始状态恢复正确的序列中，有 50 个未能完成完整变换，而显式提供状态后这 50 个案例全部得到修复。研究提出 Operational State Supervision（OSS），通过监督任务相关空间状态、变换轨迹及跨上下文共享事实对齐，尤其提升依赖状态组织与使用的 L3、L4 层级表现。
论文提出 TRAC（Trajectory-aware Reuse and Adaptive Correction，轨迹感知复用与自适应校正），这是一个无需训练的高效自回归（AR）视频生成框架。该研究发表于 arXiv:2610.02779v1，公告类型为新增论文。摘要指出，现有加速方法主要面向单轨迹生成；TRAC 旨在通过轨迹感知的复用与自适应校正提升生成效率。对开发者而言，这类无需额外训练的加速方案有望降低自回归视频生成的推理成本。
该论文研究条件图像生成中的引导扩散轨迹校正问题。作者指出，条件对齐的细微差异和视觉保真度共同决定了实际生成效果。分类器无关引导（CFG）是当前方法成功的关键，但由于缺乏显式评价准则，其优化与校正存在困难。论文提出以频谱对齐为核心的思路，旨在改进引导过程中的生成质量与条件一致性。
论文提出 SymRegFlow，一种面向视频世界模型的对称性正则化流匹配方法。原文指出，基于 Flow Matching 的多视角世界模型能够生成逼真视频，但通常受限于固定的相机机位配置。要扩展到连续变化的相机姿态，模型需要具有密集姿态覆盖范围的成对“姿态—视频”观测数据；这对数据采集和模型训练提出了更高要求。
该论文以 arXiv:2610.02626v1 形式发布，研究视觉-语言-动作（VLA）模型在机器人操作中的推理能力。论文指出，VLA 模型正越来越多地引入中间推理过程以提升机器人操作表现，但现有方法主要针对已观测状态进行推理，缺少对未来场景演化的显式预测。论文题目表明，其核心方向是通过内化时空想象来实现更高效的 VLA 推理，这可能有助于机器人在动作执行前更好地预判环境变化。
Physical AI Smart Spaces 是一个面向室内智能空间多摄像头 3D 感知的大规模基准数据集。论文作者称，据其所知，这是首个同时提供大规模、多类别和多摄像头 3D 感知数据的室内智能空间基准。数据集包含超过 280 小时的同步 1080p 视频素材。该基准有望为开发者训练、评估和比较智能空间中的多视角三维感知系统提供统一数据基础。