cs.CL（2024-10-23）

📊 共 23 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (20 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (2) 支柱七：动作重定向 (Motion Retargeting) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (20 篇)

#题目一句话要点标签🔗⭐
1 SimRAG: Self-Improving Retrieval-Augmented Generation for Adapting Large Language Models to Specialized Domains 提出SimRAG以解决专业领域RAG系统适应性问题 large language model instruction following
2 Cross-model Control: Improving Multiple Large Language Models in One-time Training 提出跨模型控制(CMC)，通过一次训练提升多个大语言模型性能。 large language model instruction following ✅
3 Multilingual Hallucination Gaps in Large Language Models 揭示大型语言模型在多语言生成中存在的幻觉差异现象 large language model
4 MiLoRA: Efficient Mixture of Low-Rank Adaptation for Large Language Models Fine-tuning MiLoRA：一种高效的混合低秩适配大语言模型微调方法 large language model
5 CogSteer: Cognition-Inspired Selective Layer Intervention for Efficiently Steering Large Language Models CogSteer：受认知启发的选择性层干预，高效引导大型语言模型 large language model
6 Beware of Calibration Data for Pruning Large Language Models 揭示校准数据对LLM剪枝的重要性，提出自生成校准数据策略。 large language model ✅
7 Large Language Models Still Exhibit Bias in Long Text LTF-TEST揭示大语言模型在长文本中仍存在偏见，FT-REGARD微调方法有效缓解偏见。 large language model
8 VoiceTextBlender: Augmenting Large Language Models with Speech Capabilities via Single-Stage Joint Speech-Text Supervised Fine-Tuning VoiceTextBlender：通过单阶段联合语音-文本监督微调增强大语言模型的语音能力 large language model
9 Towards Understanding the Fragility of Multilingual LLMs against Fine-Tuning Attacks 揭示多语言LLM在微调攻击下的脆弱性，提出安全信息定位方法。 large language model instruction following
10 Understanding When Tree of Thoughts Succeeds: Larger Models Excel in Generation, Not Discrimination 发现思维树（ToT）成功的关键：大模型擅长生成而非判别 large language model chain-of-thought
11 LEGO: Language Model Building Blocks LEGO：从大型语言模型中提取并重组小型语言模型构建块 large language model
12 CorrectionLM: Self-Corrections with SLM for Dialogue State Tracking 提出CorrectionLM，利用SLM在对话状态跟踪中实现无LLM参与的自纠正。 large language model
13 Gazelle: An Instruction Dataset for Arabic Writing Assistance Gazelle：面向阿拉伯语写作辅助的指令数据集 large language model
14 Key Algorithms for Keyphrase Generation: Instruction-Based LLMs for Russian Scientific Keyphrases 利用指令式大语言模型生成俄语科技文献关键词，提升关键词提取效果。 large language model
15 Understanding Layer Significance in LLM Alignment 提出ILA方法，揭示LLM对齐过程中各层的重要性，提升微调效率。 large language model
16 OmniFlatten: An End-to-end GPT Model for Seamless Voice Conversation OmniFlatten：一种端到端GPT模型，用于无缝语音对话 large language model
17 MojoBench: Language Modeling and Benchmarks for Mojo MojoBench：为Mojo语言建模和基准测试提供首个框架 large language model
18 Quantifying the Risks of Tool-assisted Rephrasing to Linguistic Diversity 量化工具辅助改写对语言多样性的潜在风险 large language model
19 LMLPA: Language Model Linguistic Personality Assessment LMLPA：一种评估大型语言模型语言人格的系统 large language model
20 MM-Eval: A Multilingual Meta-Evaluation Benchmark for LLM-as-a-Judge and Reward Models 提出MM-Eval多语言元评估基准，用于评估LLM作为裁判和奖励模型在多语言环境下的表现。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)

#题目一句话要点标签🔗⭐
21 Responsible Multilingual Large Language Models: A Survey of Development, Applications, and Societal Impact 构建负责任的多语言大语言模型：综述开发、应用与社会影响 curriculum learning large language model
22 Cross-lingual Transfer of Reward Models in Multilingual Alignment 研究跨语言奖励模型迁移，提升多语言对齐效果 reinforcement learning RLHF instruction following

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
23 Navigate Complex Physical Worlds via Geometrically Constrained LLM 提出基于几何约束的大语言模型，用于在复杂物理世界中进行导航。 spatial relationship large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页