cs.CL（2026-03-25）

📊 共 19 篇论文 | 🔗 3 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) (8 🔗1) 支柱五：交互与反应 (Interaction & Reaction) (1)

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

#	题目	一句话要点	标签	🔗	⭐
1	When AI Meets Early Childhood Education: Large Language Models as Assessment Teammates in Chinese Preschools	提出Interaction2Eval，利用大语言模型提升中国幼儿园师幼互动质量评估效率。	large language model
2	PoliticsBench: Benchmarking Political Values in Large Language Models with Multi-Turn Roleplay	PoliticsBench：通过多轮角色扮演评估大型语言模型中的政治价值观	large language model
3	Thinking with Tables: Enhancing Multi-Modal Tabular Understanding via Neuro-Symbolic Reasoning	提出TWT，通过神经符号推理增强表格-视觉多模态理解能力。	large language model multimodal	✅
4	How Vulnerable Are Edge LLMs?	CLIQ：针对量化边缘LLM的知识提取漏洞分析框架	large language model
5	PINGALA: Prosody-Aware Decoding for Sanskrit Poetry Generation	PINGALA：梵语诗歌生成中基于韵律感知的解码方法	large language model
6	Optimizing Multilingual LLMs via Federated Learning: A Study of Client Language Composition	通过联邦学习优化多语言LLM：客户端语言组成的影响研究	large language model
7	Alignment Reduces Expressed but Not Encoded Gender Bias: A Unified Framework and Study	提出统一框架，研究对齐训练如何影响LLM的内隐与外显性别偏见	large language model
8	ConceptKT: A Benchmark for Concept-Level Deficiency Prediction in Knowledge Tracing	ConceptKT：一个知识追踪中概念级缺陷预测的基准数据集	large language model
9	Grounding Arabic LLMs in the Doha Historical Dictionary: Retrieval-Augmented Understanding of Quran and Hadith	提出基于多哈历史词典的RAG框架，提升阿拉伯语LLM在古兰经和圣训理解上的准确性。	large language model	✅
10	OmniACBench: A Benchmark for Evaluating Context-Grounded Acoustic Control in Omni-Modal Models	OmniACBench：用于评估全模态模型中上下文相关的声学控制的基准测试。	multimodal

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#	题目	一句话要点	标签	🔗	⭐
11	GameplayQA: A Benchmarking Framework for Decision-Dense POV-Synced Multi-Video Understanding of 3D Virtual Agents	GameplayQA：提出用于评估3D虚拟智能体决策密集型第一视角多视频理解的基准框架。	world model world models embodied AI
12	Self-Distillation for Multi-Token Prediction	提出MTP-D自蒸馏方法，提升LLM多Token预测的效率和接受率。	distillation large language model
13	Representation Learning to Study Temporal Dynamics in Tutorial Scaffolding	提出基于嵌入的表征学习方法，分析辅导对话中的时间动态支架。	representation learning large language model
14	Why Does Self-Distillation (Sometimes) Degrade the Reasoning Capability of LLMs?	揭示自蒸馏对大语言模型推理能力的负面影响	distillation
15	Perturbation: A simple and efficient adversarial tracer for representation learning in language models	提出Perturbation：一种简单高效的对抗追踪器，用于语言模型中的表征学习	representation learning
16	MARCH: Multi-Agent Reinforced Self-Check for LLM Hallucination	提出MARCH，利用多智能体强化自检解决LLM幻觉问题	reinforcement learning large language model	✅
17	CoCR-RAG: Enhancing Retrieval-Augmented Generation in Web Q&A via Concept-oriented Context Reconstruction	提出CoCR-RAG，通过概念重构增强Web问答中的检索增强生成效果。	distillation large language model
18	Improving Lean4 Autoformalization via Cycle Consistency Fine-tuning	通过循环一致性微调提升Lean4自动形式化能力	reinforcement learning curriculum learning

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
19	Mechanic: Sorrifier-Driven Formal Decomposition Workflow for Automated Theorem Proving	Mechanic：基于Sorry驱动的形式化分解工作流，提升自动定理证明效率	IMoS large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页