cs.CL（2026-03-04）

📊 共 21 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) (16 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (16 篇)

#题目一句话要点标签🔗⭐
1 Position: Vector Prompt Interfaces Should Be Exposed to Enable Customization of Large Language Models 建议开放向量提示接口以实现大语言模型的可定制化 large language model
2 Traces of Social Competence in Large Language Models 通过改进的False Belief Test评估大型语言模型的社会认知能力 large language model
3 Benchmarking Motivational Interviewing Competence of Large Language Models 评估大型语言模型在动机访谈中的能力，验证其在心理咨询领域的应用潜力。 large language model
4 A Neural Topic Method Using a Large-Language-Model-in-the-Loop for Business Research LX Topic：融合大语言模型的神经主题模型，提升商业研究中文本分析质量 large language model
5 Monitoring Emergent Reward Hacking During Generation via Internal Activations 提出基于内部激活的奖励劫持监测方法，用于检测生成过程中的模型对齐问题。 large language model chain-of-thought
6 CONCUR: Benchmarking LLMs for Concurrent Code Generation CONCUR：用于评估LLM并发代码生成能力的新基准测试 large language model
7 Bielik-Q2-Sharp: A Comparative Study of Extreme 2-bit Quantization Methods for a Polish 11B Language Model Bielik-Q2-Sharp：针对波兰语11B语言模型的极端2比特量化方法对比研究 large language model
8 Rethinking Role-Playing Evaluation: Anonymous Benchmarking and a Systematic Study of Personality Effects 提出匿名评估方法，并研究人格增强对角色扮演Agent性能的影响 large language model
9 CzechTopic: A Benchmark for Zero-Shot Topic Localization in Historical Czech Documents CzechTopic：面向捷克历史文档的零样本主题定位基准 large language model ✅
10 T2S-Bench & Structure-of-Thought: Benchmarking and Prompting Comprehensive Text-to-Structure Reasoning 提出T2S-Bench基准测试和SoT提示方法，提升LLM在文本到结构推理任务上的性能。 large language model ✅
11 The Company You Keep: How LLMs Respond to Dark Triad Traits 研究大型语言模型对黑暗三角特质的响应机制 large language model
12 Retrieval or Representation? Reassessing Benchmark Gaps in Multilingual and Visually Rich RAG 重新评估多语言和视觉RAG中的基准差距：文档表示优于检索方法 multimodal
13 When Do Language Models Endorse Limitations on Human Rights Principles? 评估大型语言模型对人权原则限制的倾向与偏差 large language model
14 FINEST: Improving LLM Responses to Sensitive Topics Through Fine-Grained Evaluation FINEST：通过细粒度评估提升LLM对敏感话题的回应质量 large language model
15 Assessing the Effectiveness of LLMs in Delivering Cognitive Behavioral Therapy 评估大型语言模型在认知行为疗法中的有效性 large language model
16 ErrorLLM: Modeling SQL Errors for Text-to-SQL Refinement ErrorLLM：通过建模SQL错误来改进Text-to-SQL的生成效果 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🔗⭐
17 Confidence-Calibrated Small-Large Language Model Collaboration for Cost-Efficient Reasoning 提出COREA，通过置信度校准的大小模型协同推理，实现成本效益的复杂推理。 reinforcement learning large language model
18 World Properties without World Models: Recovering Spatial and Temporal Structure from Co-occurrence Statistics in Static Word Embeddings 静态词嵌入蕴含世界知识：无需世界模型即可恢复时空结构 world model large language model
19 Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience Memory Memex(RL)：通过索引经验记忆扩展长时程LLM Agent reinforcement learning reward shaping large language model
20 Code Fingerprints: Disentangled Attribution of LLM-Generated Code 提出解耦代码归因网络DCAN，用于识别LLM生成代码的来源模型。 contrastive learning large language model ✅
21 Linguistically Informed Graph Model and Semantic Contrastive Learning for Korean Short Text Classification LIGRAM：融合语言学知识图模型与语义对比学习的韩语短文本分类方法 contrastive learning

⬅️ 返回 cs.CL 首页 · 🏠 返回主页