cs.AI（2026-09-21）

📊 共 23 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (13) 支柱二：RL算法与架构 (RL & Architecture) (5) 支柱一：机器人控制 (Robot Control) (4) 支柱七：动作重定向 (Motion Retargeting) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (13 篇)

#	题目	一句话要点	标签	🔗	⭐
1	Representation-guided in-context learning for medical image interpretation with multimodal large language models	提出基于表示引导的上下文学习以解决医学图像解读问题	large language model multimodal
2	Construting Reverse Thinking: Developing Large Language Models' Reverse Thingking Ability	提出反向推理能力构建方法以解决复杂问题	large language model chain-of-thought
3	Graded-Relevance Composed Multimodal Retrieval for E-commerce Visual Search at Scale	提出GradCIR以解决电商视觉搜索中的分级相关性问题	multimodal
4	From Bits to Beliefs: Recoverable Semantic Fingerprints for Black-Box Verification of Large Language Models	提出SimPrint框架以解决大语言模型所有权验证问题	large language model
5	Pinocchio: Fast Uncertainty Estimates for Black-Box Language Models	提出Pinocchio以解决黑箱语言模型的不确定性估计问题	large language model zero-shot transfer
6	LADDER: Graph-Guided Diffusion Language Models for Efficient Multi-Hop Reasoning	提出LADDER以解决多跳推理效率低下问题	large language model foundation model
7	LIMIT: Less Is More for Instruction Tuning in Text-to-SQL	提出LIMIT框架以优化Text-to-SQL指令调优数据需求	large language model chain-of-thought
8	Extracting Arguments, Not Just Classifying Them: Instruction-Tuned LLMs for Generative Component Detection	提出ITFACD以解决论证成分检测问题	large language model
9	GRUET: Quantifying Uncertainty of Agentic Reasoning-and-Acting Processes	提出GRUET以量化智能体推理与行动过程的不确定性	large language model
10	Reasoning Topology Matters: A Controlled Study of LLM-Based Cybersecurity Analysis	提出安全推理拓扑以提升网络安全分析的准确性	large language model
11	TimeLitmus: A Diagnostic Benchmark for Cross-Modal Understanding and Explanation Faithfulness in Event-Conditioned Time-Series Prediction	提出TimeLitmus以解决跨模态理解与解释可信度问题	large language model
12	ActGov: Governing LLM Agent Actions via Policy-Constrained Validation	提出ActGov以解决LLM代理行动的授权管理问题	large language model
13	VLM-in-Sandbox: Visual Workspaces for Agentic Visual Reasoning	提出VLM-in-Sandbox以解决多模态视觉推理中的状态管理问题	multimodal

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#	题目	一句话要点	标签	🔗	⭐
14	LEAP-NBV: Lightweight Edge Active-Perception for Foundation-Model Next-Best-View Planning	提出LEAP-NBV以解决边缘设备上实时主动感知问题	distillation human mesh recovery HMR
15	Taming CoT Obfuscation in VLMs: From Mechanistic Evidence to Activation Enforcement	提出TAME以解决VLM中的链式思维模糊问题	reinforcement learning multimodal chain-of-thought
16	Few-Shot Demonstrations Elicit the Use of In-Context World Representations in LLMs	通过少量示例提升LLM的上下文世界表示能力	world model world models large language model
17	Harness-Zero: Harness Distillation via Agent-as-Harness	提出Harness-Zero以解决代理与环境交互的优化问题	distillation
18	Conduit: An Experience Data Plane for Distributed Reinforcement Learning	提出Conduit以优化分布式强化学习中的经验管理问题	reinforcement learning

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#	题目	一句话要点	标签	🔗	⭐
19	Decoding Guardrails: XAI-Guided Perturbation Analysis of Prompt Injection Detection	应用可解释人工智能分析提示注入检测中的防护机制	manipulation large language model
20	DeceptionAnalyser: A Web-Based AI Tool for Performing Structured Deception Analysis with Argumentation Schemes and LLMs	提出DeceptionAnalyser以解决欺骗分析的系统性挑战	manipulation large language model
21	Data Agents: Agentic Data Systems	提出数据代理以解决传统数据系统的局限性	manipulation
22	DUMA-Bench: A Dual-Control Multi-Agent Benchmark for Evaluating LLM Agent Security	提出DUMA-Bench以评估LLM代理的安全性问题	manipulation

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
23	Generative Tutorial: Towards Live Contextualized Visual Instructions for Physical Tasks	提出生成式教程以解决物理任务中的视觉指导问题	spatial relationship

⬅️ 返回 cs.AI 首页 · 🏠 返回主页