cs.CL（2024-12-23）

📊 共 28 篇论文 | 🔗 7 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (23 🔗6) 支柱二：RL算法与架构 (RL & Architecture) (3) 支柱六：视频提取与匹配 (Video Extraction) (1 🔗1) 支柱一：机器人控制 (Robot Control) (1)

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

#题目一句话要点标签🔗⭐
1 Knowledge Editing through Chain-of-Thought 提出EditCoT，通过思维链编辑实现大语言模型知识更新，无需重训练。 large language model chain-of-thought ✅
2 Unlocking Cross-Lingual Sentiment Analysis through Emoji Interpretation: A Multimodal Generative AI Approach 利用表情符号理解跨语言情感：一种多模态生成式AI方法 large language model multimodal ✅
3 A Multimodal Emotion Recognition System: Integrating Facial Expressions, Body Movement, Speech, and Spoken Language 提出一种多模态情感识别系统，融合面部表情、肢体动作、语音和语言，提升心理评估的客观性和准确性。 multimodal
4 BenCzechMark : A Czech-centric Multitask and Multimetric Benchmark for Large Language Models with Duel Scoring Mechanism 提出BenCzechMark以评估捷克语大语言模型的性能 large language model ✅
5 CARL-GT: Evaluating Causal Reasoning Capabilities of Large Language Models 提出CARL-GT基准以评估大语言模型的因果推理能力 large language model
6 Path-of-Thoughts: Extracting and Following Paths for Robust Relational Reasoning with Large Language Models 提出Path-of-Thoughts框架，解决LLM在关系推理中的难题 large language model
7 Generating Completions for Broca's Aphasic Sentences Using Large Language Models 利用大型语言模型生成补全语句，辅助改善Broca失语症患者的表达 large language model
8 A Survey of Query Optimization in Large Language Models 综述：针对大型语言模型检索增强生成中的查询优化技术 large language model
9 DRT: Deep Reasoning Translation via Long Chain-of-Thought 提出DRT：通过长链式思考进行深度推理翻译，提升隐喻和比喻句翻译质量。 chain-of-thought ✅
10 WarriorCoder: Learning from Expert Battles to Augment Code Large Language Models WarriorCoder：通过专家代码LLM对抗学习增强代码大语言模型 large language model
11 Interweaving Memories of a Siamese Large Language Model 提出IMSM框架，通过交织Siamese LLM的记忆来缓解PEFT中的灾难性遗忘问题。 large language model
12 A Dual-Perspective Metaphor Detection Framework Using Large Language Models 提出DMD双视角隐喻检测框架，利用大语言模型提升隐喻识别的透明性和可靠性。 large language model
13 Deliberation in Latent Space via Differentiable Cache Augmentation 提出基于可微缓存增强的潜在空间审议方法，提升LLM推理能力并降低延迟。 large language model
14 StructTest: Benchmarking LLMs' Reasoning through Compositional Structured Outputs 提出StructTest以解决LLMs评估中的偏差与成本问题 large language model
15 ResearchTown: Simulator of Human Research Community 提出ResearchTown，用于模拟人类研究社区，助力科学发现和创新。 large language model
16 YuLan-Mini: An Open Data-efficient Language Model YuLan-Mini：一种数据高效的开源语言模型，参数量24.2亿。 large language model ✅
17 The Power of Adaptation: Boosting In-Context Learning through Adaptive Prompting 提出Adaptive-Prompt自适应提示方法，提升大语言模型上下文学习能力 large language model
18 LiveIdeaBench: Evaluating LLMs' Divergent Thinking for Scientific Idea Generation with Minimal Context LiveIdeaBench：通过单关键词提示评估LLM在科学构思中发散性思维能力 large language model
19 A Silver Bullet or a Compromise for Full Attention? A Comprehensive Study of Gist Token-based Context Compression 研究基于Gist Token的上下文压缩方法，提升大语言模型长文本处理能力。 large language model
20 A Survey on LLM-based Multi-Agent System: Recent Advances and New Frontiers in Application 综述基于LLM的多智能体系统：最新进展与应用前沿 large language model
21 Measuring Contextual Informativeness in Child-Directed Text 提出一种基于大型语言模型的方法，用于评估儿童故事中词汇的语境信息量。 large language model
22 Just What You Desire: Constrained Timeline Summarization with Self-Reflection for Enhanced Relevance 提出约束时间线摘要（CTLS）任务，并利用自反思LLM提升摘要相关性 large language model
23 Assessing Human Editing Effort on LLM-Generated Texts via Compression-Based Edit Distance 提出基于压缩的编辑距离度量以评估人类对LLM生成文本的编辑努力 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 Diving into Self-Evolving Training for Multimodal Reasoning 提出M-STAR框架，通过自进化训练提升多模态推理性能并缓解饱和问题。 reinforcement learning multimodal chain-of-thought
25 Understanding the Logic of Direct Preference Alignment through Logic 通过逻辑形式化分析，理解并改进直接偏好对齐算法 preference learning DPO large language model
26 Brain-to-Text Benchmark '24: Lessons Learned 脑-文本转换基准'24：通过集成解码器和优化训练提升解码精度 state space model large language model

🔬 支柱六：视频提取与匹配 (Video Extraction) (1 篇)

#题目一句话要点标签🔗⭐
27 Chumor 2.0: Towards Benchmarking Chinese Humor Understanding 构建大规模中文幽默解释数据集Chumor 2.0，用于评估和提升LLM的中文幽默理解能力。 HuMoR chain-of-thought ✅

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
28 DiffusionAttacker: Diffusion-Driven Prompt Manipulation for LLM Jailbreak DiffusionAttacker：一种基于扩散模型的LLM越狱提示操控方法 manipulation large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页