cs.CL（2026-07-29）

📊 共 15 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) (10 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (10 篇)

#题目一句话要点标签🔗⭐
1 Metis: Memory Foundation Model 提出Metis以增强基础模型的原生记忆能力 foundation model multimodal
2 Dual-Path LLM Reasoning for Multimodal Few-Shot Knowledge Graph Completion 提出DuPLeR框架以解决多模态少样本知识图谱补全问题 large language model multimodal
3 Symphony of Bias: Exploring Gender Associations with Musical Instruments in Multimodal LLMs 提出Symphony-Bias数据集以探讨多模态LLMs中的性别偏见 large language model multimodal
4 From Found to Designed: Concepts as a Design Axis for Large Language Models 提出将概念作为设计轴以优化大型语言模型的结构 large language model
5 ForgetBench: Benchmarking Forgetting Dynamics of Long-Term Parametric Memory in Language Models 提出ForgetBench以系统评估语言模型的遗忘动态 large language model
6 Mergeable Model-Side Aggregation States for Long-Context Language Models 提出模型侧聚合状态以解决长上下文语言模型的聚合问题 chain-of-thought ✅
7 Evaluating Regional Bias in LLMs From Abstract Stereotype to Concrete Social Decision-Making 提出S2D框架以系统评估大型语言模型中的区域偏见 large language model
8 OptimismBench: Forecasting Bias and the Alignment Effect in Language Model Judgment 提出OptimismBench以检测语言模型判断中的偏见问题 large language model
9 Enhancing Generative Information Extraction with Two-step Validation: A Product Attribute Use Case 提出两步验证方法以提升生成信息提取的准确性 large language model
10 Revisiting Lossy Verification in Speculative Decoding: Mechanisms, Trade-offs, and Failure Modes 提出损失验证机制以提升大语言模型推理效率 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🔗⭐
11 DIRECT: Direct Decoding for Efficient and Aligned Sequence Labeling with Large Language Models 提出DIRECT框架以提升序列标注的效率与对齐性 DPO direct preference optimization large language model
12 Mental World Modeling 提出心理世界建模以解决人类决策预测问题 world model world models
13 Where Detectors Fail: Closing the Tail-Domain Gap with Expert-Guided Mutual Distillation 提出专家引导的互助蒸馏以解决多模态假新闻检测的领域偏差问题 distillation multimodal
14 SERPO: Self-Evolving Rubric Policy Optimization for Open-Ended Test-Time Reinforcement Learning 提出SERPO以解决开放式生成中的自我演化问题 reinforcement learning

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
15 AgentSnare: Learning to Delay, Divert, and Defuse Autonomous Penetration Agents 提出AgentSnare以动态欺骗渗透测试代理 penetration large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页