cs.CL（2025-12-30）

📊 共 21 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (14) 支柱二：RL算法与架构 (RL & Architecture) (7)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
1 iCLP: Large Language Model Reasoning with Implicit Cognition Latent Planning 提出iCLP框架，利用隐式认知潜在规划提升LLM推理能力 large language model chain-of-thought
2 Beyond Hallucinations: A Composite Score for Measuring Reliability in Open-Source Large Language Models 提出综合可靠性评分CRS，用于评估开源大语言模型的可靠性 large language model
3 Automated Analysis of Sustainability Reports: Using Large Language Models for the Extraction and Prediction of EU Taxonomy-Compliant KPIs 构建欧盟分类标准KPI数据集，评估大语言模型在可持续性报告分析中的能力。 large language model
4 Fantastic Reasoning Behaviors and Where to Find Them: Unsupervised Discovery of the Reasoning Process 提出RISE框架，无监督发现LLM推理过程中的行为向量并实现可控干预。 large language model chain-of-thought
5 Joint Selection for Large-Scale Pre-Training Data via Policy Gradient-based Mask Learning 提出DATAMASK，通过策略梯度优化大规模预训练数据联合选择，提升训练效率和模型性能。 large language model
6 Training a Huggingface Model on AWS Sagemaker (Without Tears) 提供Hugging Face模型在AWS SageMaker上训练的完整指南，降低云平台使用门槛。 large language model
7 Improving Multi-step RAG with Hypergraph-based Memory for Long-Context Complex Relational Modeling 提出基于超图记忆的HGMem，提升多步RAG在长文本复杂关系建模中的性能 large language model
8 LimAgents: Multi-Agent LLMs for Generating Research Limitations LimAgents：多智能体LLM框架，用于生成更深入的科研论文局限性分析。 large language model
9 Closing the Data Loop: Using OpenDataArena to Engineer Superior Training Datasets 利用OpenDataArena构建高质量训练数据集，提升大语言模型性能 large language model
10 Concept Attractors in LLMs and their Applications 利用LLM中的概念吸引子，无需训练解决翻译、幻觉等问题 large language model
11 Efficient Context Scaling with LongCat ZigZag Attention 提出LongCat ZigZag Attention，加速长文本处理，提升长程推理与Agent能力。 foundation model
12 Paragraph Segmentation Revisited: Towards a Standard Task for Structuring Speech 提出TEDPara和YTSegPara基准，并设计MiniSeg模型，用于提升语音转录文本的段落分割效果。 large language model
13 Training Report of TeleChat3-MoE TeleChat3-MoE训练报告：构建可靠高效的超大规模MoE模型训练基础设施 large language model
14 Activation Steering for Masked Diffusion Language Models 提出激活调控框架，用于控制掩码扩散语言模型生成文本的属性。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
15 Bielik 11B v3: Multilingual Large Language Model for European Languages Bielik 11B v3：面向欧洲语言的高性能多语言大语言模型 reinforcement learning DPO direct preference optimization
16 World model inspired sarcasm reasoning with large language model agents 提出WM-SAR框架，利用世界模型指导的大语言模型智能体进行反讽推理 world model large language model
17 Large Emotional World Model 提出大型情感世界模型LEWM，通过EWH数据集增强社交行为预测能力。 world model large language model
18 Detecting and Mitigating Treatment Leakage in Text-Based Causal Inference: Distillation and Sensitivity Analysis 提出文本蒸馏方法，用于检测和缓解文本因果推断中的处理泄露问题 distillation
19 Figure It Out: Improve the Frontier of Reasoning with Executable Visual States FIGR：通过可执行视觉状态提升复杂推理能力 reinforcement learning chain-of-thought
20 Disentangling Learning from Judgment: Representation Learning for Open Response Analytics 提出解耦学习与判定的表征学习框架，用于开放式作答分析，提升评分的公平性和透明度。 representation learning
21 CEC-Zero: Zero-Supervision Character Error Correction with Self-Generated Rewards CEC-Zero：基于自生成奖励的零监督中文错别字纠正 reinforcement learning PPO

⬅️ 返回 cs.CL 首页 · 🏠 返回主页