cs.CL（2025-07-02）

📊 共 31 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (26 🔗5) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (26 篇)

#题目一句话要点标签🔗⭐
1 AdamMeme: Adaptively Probe the Reasoning Capacity of Multimodal Large Language Models on Harmfulness AdamMeme：自适应探查多模态大语言模型在有害性上的推理能力 large language model multimodal ✅
2 Evaluating Large Language Models for Multimodal Simulated Ophthalmic Decision-Making in Diabetic Retinopathy and Glaucoma Screening 评估大型语言模型在糖尿病视网膜病变和青光眼筛查中多模态模拟眼科决策的能力 large language model multimodal
3 Latent Chain-of-Thought? Decoding the Depth-Recurrent Transformer 研究深度循环Transformer的潜在思维链，揭示其内部推理结构的局限性 chain-of-thought ✅
4 McBE: A Multi-task Chinese Bias Evaluation Benchmark for Large Language Models 提出McBE：一个用于评估大型语言模型中文偏见的多任务基准 large language model
5 Eka-Eval: An Evaluation Framework for Low-Resource Multilingual Large Language Models Eka-Eval：一个面向低资源多语言大语言模型的评估框架 large language model ✅
6 MuRating: A High Quality Data Selecting Approach to Multilingual Large Language Model Pretraining MuRating：一种高质量多语言大语言模型预训练数据选择方法 large language model
7 Reasoning or Not? A Comprehensive Evaluation of Reasoning LLMs for Dialogue Summarization 对话摘要任务中，推理型大语言模型表现不如非推理模型：一项综合评估研究 large language model chain-of-thought
8 Chart Question Answering from Real-World Analytical Narratives 提出基于真实分析叙事的图表问答数据集，揭示现有模型在生态有效场景下的性能差距 large language model multimodal
9 Is External Information Useful for Stance Detection with LLMs? 研究表明：外部信息通常会降低LLM在立场检测任务中的性能 large language model chain-of-thought ✅
10 How Do Vision-Language Models Process Conflicting Information Across Modalities? 研究视觉-语言模型如何处理跨模态的冲突信息，并发现可控的路由机制。 multimodal
11 The Book of Life approach: Enabling richness and scale for life course research 提出Book of Life方法，融合复杂日志数据与LLM，实现大规模、多维度的人生轨迹研究 large language model
12 Dissecting the Impact of Mobile DVFS Governors on LLM Inference Performance and Energy Efficiency 针对移动端LLM推理能效问题，提出统一的能量感知DVFS调控器FUSE。 large language model
13 AI4Research: A Survey of Artificial Intelligence for Scientific Research 构建AI4Research的全面综述，旨在促进AI在科学研究中的创新应用。 large language model
14 High-Layer Attention Pruning with Rescaling 提出高层注意力头剪枝与重缩放方法，提升LLM生成任务性能。 large language model
15 DIY-MKG: An LLM-Based Polyglot Language Learning System DIY-MKG：一个基于LLM的多语种语言学习系统，构建个性化知识图谱。 large language model
16 Low-Perplexity LLM-Generated Sequences and Where To Find Them 提出基于低困惑度序列分析的LLM训练数据溯源方法，揭示模型记忆行为 large language model
17 LLMs for Legal Subsumption in German Employment Contracts 利用LLM和上下文学习评估德国雇佣合同条款的合法性 large language model
18 Intrinsic Fingerprint of LLMs: Continue Training is NOT All You Need to Steal A Model! 提出基于注意力参数分布指纹的LLM溯源方法，可有效应对持续训练攻击。 large language model
19 Emotionally Intelligent Task-oriented Dialogue Systems: Architecture, Representation, and Optimisation 提出LUSTER：基于LLM和强化学习的情感智能任务型对话系统 large language model
20 PDFMathTranslate: Scientific Document Translation Preserving Layouts PDFMathTranslate：首个开源的科学文档翻译软件，保持版面布局。 large language model ✅
21 Efficient Out-of-Scope Detection in Dialogue Systems via Uncertainty-Driven LLM Routing 提出基于不确定性驱动的LLM路由方法，高效检测对话系统中超出范围的意图 large language model
22 Breaking Physical and Linguistic Borders: Multilingual Federated Prompt Tuning for Low-Resource Languages 提出多语言联邦Prompt Tuning，解决低资源语言场景下的数据共享和语言差异问题。 large language model
23 Symbolic or Numerical? Understanding Physics Problem Solving in Reasoning LLMs 研究推理LLM在物理问题求解中的符号推导能力，并探索Few-shot Prompting的优化潜力 large language model
24 La RoSA: Enhancing LLM Efficiency via Layerwise Rotated Sparse Activation La RoSA：通过层级旋转稀疏激活提升大语言模型效率 large language model
25 Rethinking All Evidence: Enhancing Trustworthy Retrieval-Augmented Generation via Conflict-Driven Summarization 提出CARE-RAG，通过冲突驱动的摘要增强RAG系统的可靠性，解决知识冲突问题。 large language model
26 GAIus: Combining Genai with Legal Clauses Retrieval for Knowledge-based Assistant GAIus：结合GenAI与法律条文检索的知识型助手，提升非英语国家法律咨询效果 large language model

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

#题目一句话要点标签🔗⭐
27 Gradient-Adaptive Policy Optimization: Towards Multi-Objective Alignment of Large Language Models 提出梯度自适应策略优化GAPO，解决大语言模型多目标对齐难题 reinforcement learning RLHF large language model
28 Evaluating the Effectiveness of Direct Preference Optimization for Personalizing German Automatic Text Simplifications for Persons with Intellectual Disabilities 利用直接偏好优化（DPO）个性化德语自动文本简化，提升智力障碍人士可访问性 DPO direct preference optimization large language model
29 OpenTable-R1: A Reinforcement Learning Augmented Tool Agent for Open-Domain Table Question Answering 提出OpenTable-R1，利用强化学习增强工具型Agent解决开放域表格问答问题 reinforcement learning large language model ✅
30 Skywork-Reward-V2: Scaling Preference Data Curation via Human-AI Synergy 提出Skywork-Reward-V2，通过人机协同提升偏好数据质量，显著增强奖励模型性能 reinforcement learning RLHF large language model
31 MiCoTA: Bridging the Learnability Gap with Intermediate CoT and Teacher Assistants MiCoTA：利用中间CoT和教师助手弥合小模型学习能力差距 distillation large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页