cs.AI（2026-07-09）

📊 共 20 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) (14 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (5) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
1 Concretized Proposition Prompting Resolves Composition-Knowledge Dichotomy in Large Language Models 提出具体化命题提示以解决大型语言模型的组合知识二分法问题 large language model foundation model
2 Towards Mechanistically Understanding Why Memorized Knowledge Fails to Generalize in Large Language Model Finetuning 提出自修补技术以解决大语言模型微调中的知识使用差距问题 large language model
3 Blind-Spots-Bench: Evaluating Blind Spots in Multimodal Models 提出Blind-Spots-Bench以评估多模态模型的盲点问题 multimodal
4 PARA-PV: Physics-Aware Retrieval-Augmented PV Prediction Based on Frozen Foundation Model and Distribution Shift Correction 提出PARA-PV框架以解决光伏发电预测中的物理约束问题 foundation model ✅
5 A First-Principles Theory of Slow Thinking and Active Perception 提出基于第一性原理的理论以解决慢思维与主动感知问题 large language model
6 CausalDS: Benchmarking Causal Reasoning in Data-Science Agents 提出CausalDS以解决因果推理在数据科学代理中的评估问题 large language model
7 Persuasion Attacks Can Decrease Effectiveness of CoT Monitoring 提出多模型事实核查以解决CoT监控的说服攻击问题 chain-of-thought
8 Beware What You Autocomplete: Forensic Attribution of Backdoored Code Completions 提出CodeTracer以追踪后门代码补全的来源 large language model
9 Workflow as Knowledge: Semantic Persistence for LLM-Mediated Workflows 提出语义持久性模型以优化LLM驱动的工作流 large language model
10 The Illusion of Equivalency: Statistical Characterization of Quantization Effects in LLMs 提出正确性一致性指标以评估LLM量化影响 large language model
11 Understanding Axes of Difficulty For Long Context Tasks Via PredicateLongBench 提出PredicateLongBench以评估长上下文任务的难度 large language model
12 Multi-Agent Firewall Architecture for Privacy Protection of Sensitive Data in Interactions with Language Models 提出多代理防火墙架构以保护与语言模型交互中的隐私数据 large language model
13 From Prompts to Contracts: Harness Engineering for Auditable Enterprise LLM Agents 提出一种可审计的企业LLM代理架构以解决源边界和验证问题 large language model
14 Long-Horizon-Terminal-Bench: Testing the Limits of Agents on Long-Horizon Terminal Tasks with Dense Reward-Based Grading 提出Long-Horizon-Terminal-Bench以解决长时间终端任务评估问题 multimodal

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

#题目一句话要点标签🔗⭐
15 Applying JEPA-Style Predictive Learning to JA4-Derived Network Fingerprints 提出JA4-JEPA以提升网络指纹的预测学习效果 JEPA representation learning
16 Compete Then Collaborate: Frontier AI Teachers Build a Verifiable Curriculum to Improve a Coding Student Beyond Imitation 提出竞争与协作框架以提升编程学生的学习效果 reinforcement learning distillation large language model
17 ADORN: Adaptive Drift handling for Open RAN using Reinforcement Learning 提出基于强化学习的自适应漂移处理方法以优化O-RAN性能 reinforcement learning
18 Different Teachers, Different Capabilities: Sub-1B On-Device Distillation for Structured Text Enrichment 提出基于蒸馏的模型以提升结构化文本提取效率 distillation
19 GATS: Graph-Augmented Tree Search with Layered World Models for Efficient Agent Planning 提出GATS以解决LLM代理规划中的高计算成本问题 world model world models large language model

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
20 AUTOPILOT VQA: Benchmarking Vision-Language Models for Incident-Centric Dashcam Understanding 提出AUTOPILOT-VQA以解决自动驾驶安全事件理解问题 scene understanding large language model multimodal

⬅️ 返回 cs.AI 首页 · 🏠 返回主页