cs.AI（2026-04-01）

📊 共 20 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (14 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱一：机器人控制 (Robot Control) (1) 支柱四：生成式动作 (Generative Motion) (1)

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

#题目一句话要点标签🔗⭐
1 TRACE: Training-Free Partial Audio Deepfake Detection via Embedding Trajectory Analysis of Speech Foundation Models 提出TRACE，通过分析语音基础模型的嵌入轨迹来检测无训练的部分音频深度伪造。 foundation model
2 Does Unification Come at a Cost? Uni-SafeBench: A Safety Benchmark for Unified Multimodal Large Models 提出Uni-SafeBench，评估统一多模态大模型在多任务下的安全性问题。 multimodal
3 Adversarial Moral Stress Testing of Large Language Models 提出对抗性道德压力测试框架AMST，评估LLM在对抗环境下的伦理鲁棒性。 large language model
4 OmniMem: Autoresearch-Guided Discovery of Lifelong Multimodal Agent Memory OmniMem：通过自主研究发现终身多模态Agent记忆 multimodal ✅
5 Towards Reliable Truth-Aligned Uncertainty Estimation in Large Language Models 提出真值锚定(TAC)校准方法，提升大语言模型不确定性估计的可靠性。 large language model ✅
6 HippoCamp: Benchmarking Contextual Agents on Personal Computers HippoCamp：用于评估个人电脑上上下文感知Agent的新基准 large language model multimodal
7 Automated Framework to Evaluate and Harden LLM System Instructions against Encoding Attacks 提出自动化框架评估并强化LLM系统指令，防御编码攻击 large language model chain-of-thought
8 Investigating Autonomous Agent Contributions in the Wild: Activity Patterns and Code Change over Time 研究自主编码Agent在开源项目中的贡献：活动模式与代码随时间的变化 large language model
9 Therefore I am. I Think 揭示大语言模型推理决策机制：决策先于思考，早期编码影响思维链 chain-of-thought
10 OrgAgent: Organize Your Multi-Agent System like a Company OrgAgent：构建公司式层级多智能体系统，提升复杂推理能力 large language model
11 Streaming Model Cascades for Semantic SQL 提出两种流式模型级联算法，用于在语义SQL中降低大语言模型的推理成本。 large language model
12 Ontology-Constrained Neural Reasoning in Enterprise Agentic Systems: A Neurosymbolic Architecture for Domain-Grounded AI Agents 提出本体约束神经推理框架，解决企业级Agent系统中LLM的幻觉和合规性问题 large language model
13 Adaptive Parallel Monte Carlo Tree Search for Efficient Test-time Compute Scaling 提出自适应并行蒙特卡洛树搜索，提升大模型推理时效性与吞吐量。 large language model
14 Signals: Trajectory Sampling and Triage for Agentic Interactions 提出基于信号的Agent交互轨迹采样与分流框架，提升后部署优化效率。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
15 RefineRL: Advancing Competitive Programming with Self-Refinement Reinforcement Learning 提出RefineRL，利用自精炼强化学习提升LLM在编程竞赛中的表现 reinforcement learning large language model
16 Adversarial Attacks in AI-Driven RAN Slicing: SLA Violations and Recovery 研究AI驱动的RAN切片中对抗攻击对SLA的影响与恢复策略 reinforcement learning deep reinforcement learning DRL
17 The Silicon Mirror: Dynamic Behavioral Gating for Anti-Sycophancy in LLM Agents 提出Silicon Mirror框架，通过动态行为控制减少LLM Agent中的谄媚现象 RLHF large language model
18 Execution-Verified Reinforcement Learning for Optimization Modeling 提出EVOM框架，通过执行验证强化学习实现优化建模的自动化，并支持跨求解器泛化。 reinforcement learning

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
19 Beyond Symbolic Solving: Multi Chain-of-Thought Voting for Geometric Reasoning in Large Language Models 提出MARS-GPS，通过多链思考投票提升大语言模型在几何推理上的性能。 manipulation large language model chain-of-thought

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
20 AutoEG: Exploiting Known Third-Party Vulnerabilities in Black-Box Web Applications AutoEG：利用已知第三方漏洞实现黑盒Web应用自动化渗透测试 penetration

⬅️ 返回 cs.AI 首页 · 🏠 返回主页