SkillConsist：通过双向图对齐检测 Agent Skill 不一致性
SkillConsist 提出利用双向图对齐方法检测 Agent Skill 之间的不一致性。Agent Skill 为 LLM Agent 提供可复用能力，但其不一致性可能暴露未披露的危险行为，或导致系统错误选择 Skill。随着 Agent Skill 研究的发展，一致性检测正成为提升 Agent 安全性与可靠性的重要方向。
独立 AI 资讯库
AI 前沿动态、原文归档与媒体镜像
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。
12005资讯
11873正文归档
75817媒体
多模态大语言模型不断演变的安全格局：新兴威胁与防护措施综述
该论文系统综述了多模态大语言模型（MLLM）的新兴安全威胁与防护措施。MLLM 通过模态对齐与融合整合异构数据，从而增强理解和推理能力。论文指出，这种架构转变也正在重塑机器学习的安全格局，对模型开发者和行业部署提出新的安全治理需求。
评估指标能否检测古典汉语到英语翻译中的错误？
这篇新发布的 arXiv 预印本（arXiv:2608.08283v1）探讨自动评估指标是否能够识别古典汉语译为英语时的翻译错误。研究背景是，大语言模型虽然能以令人意外的水平翻译部分历史语言，但数字人文工作流仍缺乏可靠的质量评估手段。该研究聚焦现有自动评价指标在这一特定翻译任务中的有效性，对开发者选择翻译质量评估方法、以及数字人文学者审核机器译文具有参考意义。原文提供的摘要片段未包含具体实验数据或最终结论。
Thinking vs. NoThinking: Towards Interpreting Reasoning Mechanisms of Large Language Models via Sparse Autoencoders
arXiv:2608.08168v1 Announce Type: new Abstract: While Large Language Models (LLMs) employing Chain-of-Thought (CoT) exhibit superior reasoning capabilities, the neural mechanisms distinguishing this explicit Thinking mode from direct answer generation (NoThinking mode) remain poorly understood. To
STEMMA: An Adversarial Multi-Agent Framework for Evaluating Self-Identity Consistency in LLMs
arXiv:2608.08164v1 Announce Type: new Abstract: Knowledge Distillation is a widely adopted technique in the training and fine-tuning of large language models (LLMs) enabling transfer of structured information and functional behavior from a large teacher model to a smaller student model while signif
Can LLM Agents Stick to the Script? A Benchmark for Long-Horizon Consistency in Interactive Narratives
arXiv:2608.08160v1 Announce Type: new Abstract: The rapid advancement of Large Language Models (LLMs) is revolutionizing AI for Games by enabling open-ended and fluid interactive storytelling. However, existing research has largely overlooked the critical challenge of maintaining long-horizon logic
NeuPAT: Neuron-aware Plasticity Allocation Tuning for Language-Preserving MLLMs
arXiv:2608.08107v1 Announce Type: new Abstract: Multimodal expansion of large language models (LLMs) enables new perceptual capabilities but often compromises the language intelligence acquired during pretraining. In this work, we investigate this phenomenon from the perspective of internal adaptat
Thinking Hard, Not Smart: Reasoning Models Fail to Ration Test-Time Compute Across Questions
arXiv:2608.07968v1 Announce Type: new Abstract: Reasoning language models increasingly use test-time compute to improve performance, but existing evaluations typically study this compute one question at a time. Yet when multiple problems share an end-to-end cost or latency constraint, models must d
SurakshaEval: An Indic Safety Benchmark for Multilingual LLMs
arXiv:2608.07862v1 Announce Type: new Abstract: Existing safety evaluation datasets for large language models (LLMs) predominantly focus on English and Western contexts, often overlooking the linguistic diversity and culturally grounded safety risks present in other languages. To address this gap,
SurveyReview: A Reviewer-Aligned Benchmark for Survey Evaluators
arXiv:2608.07641v1 Announce Type: new Abstract: The rapid advancement of large language models has transformed survey writing from a months-long manual effort into an automated process. As generation scales, reliable evaluation becomes the bottleneck, and LLMs are increasingly used as survey evalua
Search-G1: Grounded Search Agents via Representation-Based Intrinsic Rewards
arXiv:2608.07531v1 Announce Type: new Abstract: Search-augmented language agents should retrieve external information only when necessary and ground their answers in retrieved evidence. Existing external rewards provide either sparse outcome supervision or richer feedback from process annotations a
WuYuEval: A Multi-Level Benchmark for Large Language Models in Solid Waste Management
arXiv:2608.07529v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as technical assistants, but their competence in solid waste management (SWM) remains difficult to assess because existing benchmarks emphasize general knowledge rather than professional decisions und
DocAtlas: Long-Document Understanding as Mutable-State Interaction
arXiv:2608.07527v1 Announce Type: new Abstract: Long-document understanding requires models to find and combine evidence across many pages, layouts, tables, figures, and charts. Existing retrieval-augmented systems usually select evidence from a static index before generation, while recent agentic
Unified Hallucination Fuzzing for Multimodal Large Language Models
arXiv:2608.07525v1 Announce Type: new Abstract: Hallucination remains a persistent challenge for Multimodal Large Language Models (MLLMs), severely limiting their reliability in high-stakes applications. Existing evaluations, predominantly based on static benchmarks, suffer from narrow taxonomical
AndroidReality: How Far Are Mobile Agents from the Real World?
arXiv:2608.07775v1 Announce Type: new Abstract: Mobile agents have achieved promising results on clean online benchmarks such as AndroidWorld, yet their performance often degrades sharply in real-world deployment due to environmental variations and imperfect interface conditions. In this work, we i
Who Verifies the Benchmark? Decentralizing Trust in Large Language Model Evaluation
arXiv:2608.07762v1 Announce Type: new Abstract: LLM benchmarks can build an organization's reputation and attract customers, but only when results are transparent and verifiable. Unverified claims that DeepSeek R1 outperformed OpenAI's o1 contributed to market panic on January 27, 2025, when Nvidia
QuantumMind: Constraint-Grounded Agentic Reasoning for Speedup Analysis in Quantum Computing
arXiv:2608.07743v1 Announce Type: new Abstract: Identifying a meaningful quantum speedup requires more than matching a classical problem to a familiar quantum primitive: the claim must preserve the task, respect access and output models, expose required promises, and remain within a defensible comp
Towards Researcher Agents for Knowledge-Graph Question Answering
arXiv:2608.07700v1 Announce Type: new Abstract: Translating a natural-language question into a SPARQL query that can be executed against a large knowledge graph requires resolving lexical ambiguity, grounding surface terms in the target ontology, and producing graph patterns that are both syntactic
IntelliAudit: Using Large Language Models to Evaluate Audit Controls
arXiv:2608.07688v1 Announce Type: new Abstract: IT audits require auditors to judge whether heterogeneous organizational evidence satisfies semantic security and compliance controls. This judgment is difficult to automate because relevant evidence is distributed across policies, records, spreadshee
An Agentic AI Framework Overcomes Fundamental Limitations of Large Language Models for Glaucoma Detection from Fundus Photography
arXiv:2608.07651v1 Announce Type: new Abstract: Large language models (LLMs) show promise in medical image interpretation but suffer from hallucination, limited accuracy, and run-to-run inconsistency. We developed and validated an agentic AI framework integrating LLMs with specialized deep learning
Mendel G\"odel Machine: Recursive Self-Improving Coding Agents via Comparative Evolution
arXiv:2608.07645v1 Announce Type: new Abstract: Self-improving coding agents that iteratively rewrite their own source code have demonstrated impressive performance on coding tasks. However, existing solutions generally derive self-modification from a single failure trajectory at a time, overlookin
Contextual Value Alignment via Multilayer Combinatorial Fusion
arXiv:2608.07642v1 Announce Type: new Abstract: Aligning large language models (LLMs) with human values remains a major challenge, especially for trustworthy AI. While existing approaches such as RLHF, CAI, and their variants have achieved promising results, they often rely on a single-agent framew
Agent-MD: Selective LLM Intervention with Event-Driven Escalation for Stateful GCMC--MD Campaigns
arXiv:2608.07637v1 Announce Type: new Abstract: Long-running molecular simulation campaigns require repeated continuation from saved states, provenance-aware progression, adaptive assessment, and occasional interpretation of workflow conditions that cannot be resolved safely by fixed rules. Here, w
From Single Chatbots to Governed Agent Ecosystems: An Agentic AI Pattern Catalogue and Orchestration Framework for Mission-Critical Hospital Information Management Systems
arXiv:2608.07627v1 Announce Type: new Abstract: Hospitals are racing to embed AI, while coping with the surge in adaptation of the technology in other industries, into the triage management, documentation, scheduling, and revenue-cycle workflows, yet most deployments remain as fragmented pilots tha
Controlled Memory Interference in Continual LLM Agents
arXiv:2608.07622v1 Announce Type: new Abstract: Long-term memory enables AI agents to maintain continuity across sessions, personalize behavior, and evolve through accumulated experience. Yet memory evolution is not simply a process of storing more information: new experiences may reinforce, revise
CMU-Drive and V2V-VLA: Cooperative Multi-agent Unified Driving with Reasoning Benchmark and Vehicle-to-Vehicle Vision-Language-Action Models
arXiv:2608.07621v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have recently achieved impressive performance for end-to-end autonomous driving, yet existing approaches are primarily designed for an individual single autonomous driving agent with limited support for cooperative
TeXFix-Bench: An Empirically Grounded Multi-Format Benchmark for LLM-Based Document Source Repair
arXiv:2608.07617v1 Announce Type: new Abstract: Scientific and technical writing depends on markup sources that must compile: LaTeX, Typst, and Markdown pipelines fail on missing delimiters, mismatched environments, broken imports, or package conflicts. Existing document-repair evaluations inject f
TREAT: Evaluating Access to Formal Knowledge across Equivalent Mathematical Representations
arXiv:2608.07540v1 Announce Type: new Abstract: AI systems increasingly operate between flexible input representations and formal objects used by downstream tools. A key challenge is recognizing when an unfamiliar formulation denotes a known formal object. We study this challenge through theorem re
When LLM Agents Negotiate: Private Information and Dynamic Bargaining in Supply Chains
arXiv:2608.07538v1 Announce Type: new Abstract: As LLM agents move from decision support to autonomous procurement, firms need to know whether delegated negotiators create value, divide it predictably, and avoid money-losing contracts. We study this in a canonical supply chain bargaining problem: a
MetaSpace: Metamorphic Testing for Spatial Cognition in Embodied Agents
arXiv:2608.07533v1 Announce Type: new Abstract: An embodied agent is an intelligent entity that interacts with its environment through a physical body. Currently, the evaluation of embodied agents primarily relies on two paradigms: (1) manually annotated Visual Question Answering (VQA) pairs and (2