cs.CL（2026-01-23）

📊 共 27 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) (23 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗1) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (23 篇)

#题目一句话要点标签🔗⭐
1 Learning Domain Knowledge in Multimodal Large Language Models through Reinforcement Fine-Tuning 提出基于强化微调的多模态大语言模型领域知识学习方法 large language model multimodal
2 AuroraEdge-V-2B: A Faster And Stronger Edge Visual Large Language Model 提出AuroraEdge-V-2B，一种快速、强大的边缘视觉大语言模型，加速工业应用部署。 large language model multimodal
3 Trapped in the past? Disentangling fluid and crystallized intelligence of large language models using chess 利用国际象棋解耦大语言模型的流体智力和晶体智力 large language model
4 Standardizing Longitudinal Radiology Report Evaluation via Large Language Model Annotation 提出基于大型语言模型的放射报告纵向信息自动标注流水线，提升报告评估标准化水平。 large language model
5 Cite-While-You-Generate: Training-Free Evidence Attribution for Multimodal Clinical Summarization 提出一种免训练的证据溯源框架，用于多模态临床摘要生成。 multimodal
6 Strategies for Span Labeling with Large Language Models 针对LLM的Span标注，提出LogitMatch约束解码方法，提升匹配精度。 large language model
7 Large Language Models as Automatic Annotators and Annotation Adjudicators for Fine-Grained Opinion Analysis 利用大语言模型作为自动标注器和仲裁器，解决细粒度情感分析的数据标注难题。 large language model
8 Exploring the Effects of Alignment on Numerical Bias in Large Language Models 研究对齐方式对大语言模型数值偏差的影响，并提出缓解策略。 large language model
9 Persuasion Tokens for Editing Factual Knowledge in LLMs 提出说服令牌(P-Tokens)，实现LLM中高效的事实知识编辑，无需特定示例。 large language model
10 Curate-Train-Refine: A Closed-Loop Agentic Framework for Zero Shot Classification 提出Curate-Train-Refine框架，利用LLM动态生成监督信号进行零样本分类 large language model
11 White-Box Sensitivity Auditing with Steering Vectors 提出基于激活向量调控的白盒敏感性审计框架，用于评估LLM中的潜在偏见。 large language model ✅
12 PLawBench: A Rubric-Based Benchmark for Evaluating LLMs in Real-World Legal Practice PLawBench：一个基于规则的真实法律实践LLM评估基准 large language model ✅
13 Attention-MoA: Enhancing Mixture-of-Agents via Inter-Agent Semantic Attention and Deep Residual Synthesis 提出Attention-MoA，通过Agent间语义注意力机制增强混合Agent模型性能。 large language model
14 LOGICAL-COMMONSENSEQA: A Benchmark for Logical Commonsense Reasoning 提出LOGICAL-COMMONSENSEQA基准，用于评估常识推理中的逻辑组合能力。 chain-of-thought
15 Jacobian Scopes: token-level causal attributions in LLMs 提出Jacobian Scopes，用于量化LLM中token级别因果归因，揭示模型预测的关键影响因素。 large language model ✅
16 Cross-Lingual Activation Steering for Multilingual Language Models 提出跨语言激活调控(CLAS)方法，提升多语言模型在低资源语言上的性能。 large language model
17 Select or Project? Evaluating Lower-dimensional Vectors for LLM Training Data Explanations 提出基于架构信息的梯度选择方法，提升LLM训练数据解释的效率与准确性 large language model
18 MultiLexNorm++: A Unified Benchmark and a Generative Model for Lexical Normalization for Asian Languages 提出MultiLexNorm++基准和基于LLM的生成模型，用于亚洲语言词汇归一化 large language model
19 How Does Personalized Memory Shape LLM Behavior? Benchmarking Rational Preference Utilization in Personalized Assistants 提出RPEval基准与RP-Reasoner模型，解决个性化LLM助手中的非理性记忆利用问题 large language model ✅
20 PROST-LLM: Progressively Enhancing the Speech-to-Speech Translation Capability in LLMs PROST-LLM：渐进式提升大语言模型语音到语音翻译能力 large language model
21 Retrieve-Refine-Calibrate: A Framework for Complex Claim Fact-Checking 提出Retrieve-Refine-Calibrate框架，提升复杂声明事实核查的准确性。 large language model
22 SearchLLM: Detecting LLM Paraphrased Text by Measuring the Similarity with Regeneration of the Candidate Source via Search Engine 提出SearchLLM，通过搜索引擎辅助检测LLM生成的复述文本。 large language model
23 TL-GRPO: Turn-Level RL for Reasoning-Guided Iterative Optimization 提出Turn-Level GRPO算法，用于解决迭代优化任务中turn级别的精细化优化问题。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
24 Mixing Expert Knowledge: Bring Human Thoughts Back To the Game of Go LoGos：融合专家知识，提升LLM在围棋领域的推理能力 reinforcement learning large language model chain-of-thought ✅
25 Timely Machine: Awareness of Time Makes Test-Time Scaling Agentic 提出Timely Machine，通过时间感知的测试时缩放提升Agent能力 reinforcement learning large language model
26 Mitigating Bias in Automated Grading Systems for ESL Learners: A Contrastive Learning Approach 提出基于对比学习的自动评分系统偏见缓解方法，提升ESL学习者评分公平性 contrastive learning

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
27 Persona Jailbreaking in Large Language Models 提出PHISH框架，揭示并利用LLM中人格易受对话历史操纵的漏洞 manipulation large language model ✅

⬅️ 返回 cs.CL 首页 · 🏠 返回主页