a

arXiv cs.AI热榜

更新于刚刚 · 50 条热点

10月3日 11:20 更新，当前榜单前三是「Heavy-Tailed Memory…」。「When Do Causal Worl…」。「From Proposal to Ve…」。这是一个人工智能方向的信息源，内容以模型进展、研究与工具发布为主。本页当前收录 50 条，保留arXiv cs.AI的原始排序和指标，不把不同平台的数据直接相加，每条标题尽量直达原始页面。热摸爽只做聚合与排序，内容版权归arXiv cs.AI及原作者所有。

也常被搜索为：arXiv cs.AI热点

Heavy-Tailed Memory Traces in Long-Horizon Language Agents

arXiv cs.AI 3790.5 万热度发布于 23小时前抓取于刚刚

When Do Causal World Models Help Modular LLM Agents

arXiv cs.AI 3622.4 万热度发布于 23小时前抓取于刚刚

From Proposal to Verified Effect: Praxa, an Evidence-Bound Harness for Governed AI Agent Execution

arXiv cs.AI 3458.5 万热度发布于 23小时前抓取于刚刚

What Do Rationales Communicate? A Message-Intervention Study in Role-Specialized QA

arXiv cs.AI 3298.7 万热度发布于 23小时前抓取于刚刚

Measuring the Microtask Eligibility Gap: When Is an Off-the-Shelf SLM Enough for an Agent Harness?

arXiv cs.AI 3142.9 万热度发布于 23小时前抓取于刚刚

Characterizing a Configuration Where Inference-Time PRM-Pruned Fragment Grafting Is Inert: Evidence from Three Reasoning LMs

arXiv cs.AI 2991.3 万热度发布于 23小时前抓取于刚刚

Gradient-Aligned Pair Selection for Personalized Preference Optimization

arXiv cs.AI 2843.8 万热度发布于 23小时前抓取于刚刚

K-Dense BYOK: An Open-Source AI Research Assistant That Runs Locally and Keeps a Hash-Chained Lab Notebook

arXiv cs.AI 2700.3 万热度发布于 23小时前抓取于刚刚

Scientific Agents: Evaluating Profession-Specific System Prompts on Scientific Tasks

arXiv cs.AI 2560.9 万热度发布于 23小时前抓取于刚刚

Comedic Fool's Gold: Reward Exploits and Countermeasures in Conversational Humor

arXiv cs.AI 2425.5 万热度发布于 23小时前抓取于刚刚

EviGraph: Proof-Carrying Selective Recommendation over Temporal Public-Service Knowledge Graphs

arXiv cs.AI 2294.1 万热度发布于 23小时前抓取于刚刚

Build2SPARQL: A Large-Scale Text-to-SPARQL Benchmark Dataset for Building Knowledge Graph Querying

arXiv cs.AI 2166.7 万热度发布于 23小时前抓取于刚刚

Robust Is Salient: An Informed Adversary Moves the Optimal Signal onto the Salience Pole

arXiv cs.AI 2043.2 万热度发布于 23小时前抓取于刚刚

Conflicting Supervision Moves Commitment, Not Capability: A 12.29{\sigma} arrangement effect that is exactly zero under a convention-agnostic score

arXiv cs.AI 1923.7 万热度发布于 23小时前抓取于刚刚

Knowing When to Yield: Grounded Arbitration of User Corrections in Text-Based Embodied Agents

arXiv cs.AI 1808.1 万热度发布于 23小时前抓取于刚刚

Rules to Tools: Executable Checks for LLM Agents in Scientific Computing

arXiv cs.AI 1696.4 万热度发布于 23小时前抓取于刚刚

Predictive Credit: Measuring What Scientific Explanations Add to Experimental Forecasts

arXiv cs.AI 1588.6 万热度发布于 23小时前抓取于刚刚

ContractRL: Shielded Group-Relative Policy Optimization for Auditable Tool-Call Repair

arXiv cs.AI 1484.6 万热度发布于 23小时前抓取于刚刚

Mathematical Transfer in LLMs Follows Reasoning Approach More Than Topic

arXiv cs.AI 1384.4 万热度发布于 23小时前抓取于刚刚

Fault-Tolerant Budget Conservation in Distributed Multi-Agent Delegation

arXiv cs.AI 1288.1 万热度发布于 23小时前抓取于刚刚