cs.AI（2026-10-05）

📊 共 26 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (19 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (6) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (19 篇)

#题目一句话要点标签🔗⭐
1 Visual Grounding Safety in Vision-Language Models 提出视觉基础安全性方法以解决视觉语言模型的安全问题 visual grounding
2 Large Language Model-Guided Discovery of Weight-Five Bivariate Bicycle Codes 利用大型语言模型发现五重二元自行车码 large language model
3 Normality Constraint Learning: Adapting Foundation Models for Time Series Anomaly Detection 提出正常性约束学习以解决时间序列异常检测问题 foundation model
4 Multimodal Safety Evaluation Should Measure Controllability Beyond Classification 提出多模态安全评估方法以超越分类测量可控性 multimodal
5 CVIF: A Criticality-Driven Visual Intervention Framework for Geometric Diagram Understanding in MLLMs 提出CVIF框架以解决几何图表理解问题 large language model multimodal
6 BazaarBench: Delegation Safety in Decentralized C2C Marketplaces Run by LLM Agents 提出BazaarBench以评估去中心化C2C市场中LLM代理的安全性 large language model
7 ANT: A Multi-Granularity Network Traffic Dataset and Benchmark for Agents Behavior Auditing 提出ANT数据集以解决代理行为审计中的网络流量分析问题 large language model
8 Back to the Future: Rethinking EDA Infrastructure for Agentic Systems in Chip Design Verification 提出BTTF框架以解决芯片设计验证中的基础设施不足问题 large language model
9 FREA: A Multi-Source Expert Benchmark for Reaction Feasibility Verification 提出FREA基准以验证化学反应的可行性 large language model
10 HERA: Harness-Environment Co-Evolution for Reliable Agentic Abstention 提出HERA框架以解决智能体自我放弃问题 large language model
11 GraphDecide: Benchmarking System One Models on Graph Tasks 提出GraphDecide基准以评估系统一模型在图任务中的表现 large language model ✅
12 ImproveAnyTask: An Autonomous Post-Training Harness for Iterative Model Self-Improvement 提出ImproveAnyTask以解决模型自我改进的挑战 large language model
13 Artificial Intelligence and the New Science of Culture 提出一种新框架以测量认知与文化的共构关系 multimodal
14 AI-Decision Checkpoints for AI-Augmented Business Process Management: Framework and Educational Instantiation 提出AI决策检查点以解决BPM中AI应用不足问题 large language model
15 Copies or Sources? Measuring How LLM Aggregators Count Restated Evidence in Multi-Agent Systems 提出一种新方法以优化多智能体系统中LLM聚合器的证据计数 large language model
16 Benchmarking Jailbreak Guardrails for Embodied Agents 提出系统评估框架以优化具身智能体的安全防护 large language model
17 AgentSpy: Making AI Agent Behavior Observable 提出AgentSpy以解决AI代理行为不可观察问题 large language model
18 Beyond Semantic Similarity: Performance and Costs of Agentic Retrieval for Complex Tasks 提出代理检索方法以解决复杂任务中的信息检索问题 large language model
19 How Should a Prompt Optimizer Spend a Tight Budget? BudgetAPO with Noise-Adaptive Evaluation 提出BudgetAPO以解决紧预算下的自动提示优化问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
20 Mind the Execution Gap: Action-Semantic Mismatch in World-Model Control 提出执行一致性接口以解决世界模型控制中的动作语义不匹配问题 world model world models dreamer
21 GPlaceRL: An Open-Source Graph Reinforcement Learning Framework for Detailed Placement 提出GPlaceRL框架以解决详细布局优化问题 reinforcement learning PPO
22 RAISED: Self-Distillation for Robustness to Prompt Injection in LLM Agents 提出RAISED以解决大语言模型代理的提示注入问题 distillation
23 You Changed Your Mind, The Model Didn't: Demystifying Intent in Multi-Turn Dialogue 提出Intent-OPSD框架以解决多轮对话中的意图混淆问题 distillation large language model
24 FreSia: Frequency-Semantic Instantiation and Alignment for Multivariate Time Series Analysis 提出FreSia以解决多变量时间序列分析中的结构感知问题 MAE large language model
25 Grounded Joint-Attention Other-Play for Zero-Shot Coordination 提出MATE以解决零-shot协调问题 reinforcement learning mutual attention

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
26 Runaway Reaction: When Benign Skills Compose into Malicious Behavior 提出CRIME以解决技能组合引发的恶意行为问题 MPC

⬅️ 返回 cs.AI 首页 · 🏠 返回主页