cs.CL（2025-09-08）

📊 共 24 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (18 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (18 篇)

#题目一句话要点标签🔗⭐
1 Towards EnergyGPT: A Large Language Model Specialized for the Energy Sector 提出EnergyGPT，一个面向能源领域的专业大型语言模型，通过微调LLaMA 3.1-8B实现。 large language model
2 Toward Purpose-oriented Topic Model Evaluation enabled by Large Language Models 提出一种基于大语言模型的面向目的的动态主题模型自动评估框架 large language model ✅
3 MedBench-IT: A Comprehensive Benchmark for Evaluating Large Language Models on Italian Medical Entrance Examinations MedBench-IT：首个意大利医学入学考试LLM综合评测基准 large language model
4 EPT Benchmark: Evaluation of Persian Trustworthiness in Large Language Models 提出EPT基准，评估大型语言模型在波斯语环境下的可信度 large language model ✅
5 A Comparative Benchmark of Large Language Models for Labelling Wind Turbine Maintenance Logs 提出风机维护日志标注的LLM基准测试框架，助力运维数据分析 large language model
6 HAVE: Head-Adaptive Gating and ValuE Calibration for Hallucination Mitigation in Large Language Models HAVE：通过头自适应门控与值校准缓解大语言模型幻觉 large language model
7 LLM Analysis of 150+ years of German Parliamentary Debates on Migration Reveals Shift from Post-War Solidarity to Anti-Solidarity in the Last Decade 利用LLM分析德国议会百年辩论，揭示从战后团结到反团结的转变 large language model
8 Neurocognitive Modeling for Text Generation: Deep Learning Architecture for EEG Data 提出基于RNN编码器和Gemma 2B的分类器-LLM架构，用于脑电信号文本生成。 large language model
9 On the Same Wavelength? Evaluating Pragmatic Reasoning in Language Models across Broad Concepts 提出评估框架以提升语言模型的实用推理能力 chain-of-thought
10 Proof-Carrying Numbers (PCN): A Protocol for Trustworthy Numeric Answers from LLMs via Claim Verification 提出Proof-Carrying Numbers以解决大型语言模型的数字可信性问题 large language model
11 COMPACT: Common-token Optimized Model Pruning Across Channels and Tokens 提出COMPACT，通过联合优化词表和FFN通道剪枝，提升LLM和SLM的效率。 large language model
12 Saturation-Driven Dataset Generation for LLM Mathematical Reasoning in the TPTP Ecosystem 利用TPTP生态系统和饱和驱动的数据集生成方法，提升LLM的数学推理能力。 large language model ✅
13 MoGU V2: Toward a Higher Pareto Frontier Between Model Usability and Security MoGU V2：提升LLM可用性与安全性帕累托前沿，解决安全与可用性trade-off问题 large language model
14 Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint 提出ProCon方法，通过投影约束缓解指令微调中大语言模型的安全性风险。 large language model
15 Guided Decoding and Its Critical Role in Retrieval-Augmented Generation 研究引导解码在检索增强生成中的作用，提升输出质量并减少幻觉 large language model
16 How Small Transformation Expose the Weakness of Semantic Similarity Measures 揭示语义相似度度量方法的弱点：小变换带来的挑战 large language model
17 LAMDAS: LLM as an Implicit Classifier for Domain-specific Data Selection LAMDAS：利用LLM作为隐式分类器进行领域数据选择 large language model
18 Do LLMs exhibit the same commonsense capabilities across languages? MULTICOM基准测试揭示LLM在多语言常识生成能力上的显著差距 large language model ✅

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
19 Revolutionizing Reinforcement Learning Framework for Diffusion Large Language Models TraceRL：面向扩散语言模型的轨迹感知强化学习框架，提升推理性能。 reinforcement learning curriculum learning large language model ✅
20 IntrEx: A Dataset for Modeling Engagement in Educational Conversations IntrEx：构建教育对话中兴趣建模的大型数据集，提升LLM在教育场景的对话能力 reinforcement learning RLHF teacher-student
21 SLiNT: Structure-aware Language Model with Injection and Contrastive Training for Knowledge Graph Completion SLiNT：通过注入和对比训练的结构感知语言模型，用于知识图谱补全 representation learning contrastive learning large language model
22 Beyond Two-Stage Training: Cooperative SFT and RL for LLM Reasoning 提出协同SFT与RL训练框架，解决LLM推理中灾难性遗忘与效率问题 reinforcement learning large language model
23 The Majority is not always right: RL training for solution aggregation 提出AggLM，通过强化学习训练聚合器，提升LLM在推理任务中的表现。 reinforcement learning large language model
24 WebExplorer: Explore and Evolve for Training Long-Horizon Web Agents WebExplorer：通过探索和演化训练长程Web代理 reinforcement learning large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页