cs.CL（2025-03-05）

📊 共 34 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) (29 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (1) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (29 篇)

#题目一句话要点标签🔗⭐
1 Attentive Reasoning Queries: A Systematic Method for Optimizing Instruction-Following in Large Language Models 提出ARQs，通过领域专家知识引导LLM，显著提升复杂指令跟随能力 large language model instruction following chain-of-thought
2 MA-LoT: Model-Collaboration Lean-based Long Chain-of-Thought Reasoning enhances Formal Theorem Proving MA-LoT：模型协作的长链式思考提升Lean形式化定理证明 large language model chain-of-thought
3 The Devil Is in the Details: Tackling Unimodal Spurious Correlations for Generalizable Multimodal Reward Models 提出Shortcut-aware算法，解决多模态奖励模型中的单模态虚假相关性问题 large language model multimodal
4 Psy-Copilot: Visual Chain of Thought for Counseling 提出Psy-Copilot，利用视觉化CoT辅助心理咨询，提升LLM可解释性 large language model chain-of-thought
5 Analogical Reasoning Inside Large Language Models: Concept Vectors and the Limits of Abstraction 探究大语言模型中的类比推理：概念向量与抽象能力的局限性 large language model
6 Open-Source Large Language Models as Multilingual Crowdworkers: Synthesizing Open-Domain Dialogues in Several Languages With No Examples in Targets and No Machine Translation 利用开源大语言模型作为多语众包工人，零样本合成多语言开放域对话 large language model
7 Enhancing Collective Intelligence in Large Language Models Through Emotional Integration 通过情感集成增强大语言模型的集体智能 large language model
8 Taxation Perspectives from Large Language Models: A Case Study on Additional Tax Penalties 提出PLAT基准，评估大语言模型在额外税收处罚合法性预测中的能力 large language model
9 "Only ChatGPT gets me": An Empirical Analysis of GPT versus other Large Language Models for Emotion Detection in Text 评估大型语言模型在文本情感检测中的能力，重点对比ChatGPT与其他LLM。 large language model
10 Performance Comparison of Large Language Models on Advanced Calculus Problems 对比七种大型语言模型在高等微积分问题上的性能，揭示其优势与不足。 large language model
11 Developing and Utilizing a Large-Scale Cantonese Dataset for Multi-Tasking in Large Language Models 构建大规模粤语数据集，提升大语言模型在粤语多任务处理上的性能 large language model
12 Token-Level Privacy in Large Language Models 提出dchi-stencil以解决大语言模型中的隐私问题 large language model
13 Visualising Policy-Reward Interplay to Inform Zeroth-Order Preference Optimisation of Large Language Models 提出ZOPrO：一种用于大型语言模型偏好优化的零阶优化算法。 large language model ✅
14 iNews: A Multimodal Dataset for Modeling Personalized Affective Responses to News iNews：一个用于建模个性化情感反应的大规模多模态新闻数据集 multimodal
15 DSVD: Dynamic Self-Verify Decoding for Faithful Generation in Large Language Models 提出DSVD：一种动态自验证解码框架，用于提升大语言模型生成内容的可靠性。 large language model
16 PowerAttention: Exponentially Scaling of Receptive Fields for Effective Sparse Attention 提出PowerAttention以解决长上下文处理中的稀疏注意力问题 large language model
17 Monitoring Decoding: Mitigating Hallucination via Evaluating the Factuality of Partial Response during Generation 提出Monitoring Decoding框架，通过动态监测和干预生成过程，缓解大语言模型中的幻觉问题。 large language model
18 Process-based Self-Rewarding Language Models 提出基于过程的自奖励语言模型，提升数学推理能力。 large language model
19 Replicating Human Social Perception in Generative AI: Evaluating the Valence-Dominance Model 评估Valence-Dominance模型，研究生成AI在人类社会感知上的复现能力 multimodal
20 Framing the Game: How Context Shapes LLM Decision-Making 提出情境框架评估方法，揭示LLM决策对上下文的敏感性 large language model
21 Extrapolation Merging: Keep Improving With Extrapolation and Merging 提出Extrapolation Merging，无需额外计算资源和数据即可持续提升LLM性能。 large language model
22 Geometry-Guided Adversarial Prompt Detection via Curvature and Local Intrinsic Dimension 提出CurvaLID，利用几何特性高效检测大语言模型中的对抗性提示 large language model
23 RASD: Retrieval-Augmented Speculative Decoding 提出RASD：检索增强的推测解码加速LLM推理，提升领域外泛化性。 large language model
24 Cite Before You Speak: Enhancing Context-Response Grounding in E-commerce Conversational LLM-Agents 提出一种电商对话LLM-Agent的引用生成方法，提升事实 grounding 和用户信任度。 large language model
25 Towards Robust Universal Information Extraction: Benchmark, Evaluation, and Solution 提出RUIE-Bench以解决通用信息提取的鲁棒性问题 large language model
26 Psy-Insight: Explainable Multi-turn Bilingual Dataset for Mental Health Counseling 构建Psy-Insight：一个面向心理健康咨询的可解释多轮双语数据集 large language model
27 Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders 利用稀疏自编码器提取特征，提升人工智能文本检测的可解释性 large language model
28 EnigmaToM: Improve LLMs' Theory-of-Mind Reasoning Capabilities with Neural Knowledge Base of Entity States 提出EnigmaToM框架，利用神经知识库提升LLM的心智理论推理能力 large language model
29 SEOE: A Scalable and Reliable Semantic Evaluation Framework for Open Domain Event Detection 提出SEOE框架，通过语义评估和可扩展基准解决开放域事件检测的评估难题。 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🔗⭐
30 External Reliable Information-enhanced Multimodal Contrastive Learning for Fake News Detection 提出ERIC-FND框架，利用外部可靠信息增强多模态对比学习，提升假新闻检测性能。 contrastive learning multimodal
31 Improving LLM Safety Alignment with Dual-Objective Optimization 提出双目标优化方法DOOR，提升LLM抵抗对抗攻击的安全对齐能力 DPO direct preference optimization large language model ✅
32 Improving Neutral Point-of-View Generation with Data- and Parameter-Efficient RL 提出参数高效强化学习(PE-RL)方法，提升大语言模型生成中立观点回复的能力。 reinforcement learning RLHF large language model

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
33 Vision-Language Models Struggle to Align Entities across Modalities 提出MATE基准，揭示视觉-语言模型在跨模态实体对齐上的不足 scene understanding multimodal chain-of-thought

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
34 Enhancing Spoken Discourse Modeling in Language Models Using Gestural Cues 提出融合手势信息的语言模型，提升口语对话建模中关键线索的预测精度。 VQ-VAE

⬅️ 返回 cs.CL 首页 · 🏠 返回主页