cs.CL（2025-08-06）

📊 共 41 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (32 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (8) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (32 篇)

#题目一句话要点标签🔗⭐
1 Chain of Questions: Guiding Multimodal Curiosity in Language Models 提出Chain of Questions框架以增强多模态语言模型的推理能力 large language model foundation model multimodal
2 Training-Free Multimodal Large Language Model Orchestration 提出多模态大语言模型编排以解决模型集成问题 large language model multimodal
3 DP-GPT4MTS: Dual-Prompt Large Language Model for Textual-Numerical Time Series Forecasting 提出DP-GPT4MTS以解决文本与数值时间序列预测问题 large language model multimodal
4 GM-PRM: A Generative Multimodal Process Reward Model for Multimodal Mathematical Reasoning 提出GM-PRM以解决多模态数学推理中的错误纠正问题 large language model multimodal
5 Beyond the Leaderboard: Rethinking Medical Benchmarks for Large Language Models 提出MedCheck框架以解决医疗基准评估的可靠性问题 large language model
6 Do Recommender Systems Really Leverage Multimodal Content? A Comprehensive Analysis on Multimodal Representations for Recommendation 利用大型视觉语言模型提升多模态推荐系统的表现 multimodal
7 Improving Crash Data Quality with Large Language Models: Evidence from Secondary Crash Narratives in Kentucky 利用大型语言模型提升肯塔基州事故数据质量 large language model
8 Hierarchical Text Classification Using Black Box Large Language Models 利用黑箱大语言模型进行层次文本分类以应对数据稀缺问题 large language model
9 Efficient Strategy for Improving Large Language Model (LLM) Capabilities 提出高效策略以提升大型语言模型能力 large language model
10 A Few Words Can Distort Graphs: Knowledge Poisoning Attacks on Graph-based Retrieval-Augmented Generation of Large Language Models 提出知识毒化攻击以解决GraphRAG模型的安全隐患问题 large language model
11 Enhancing Dialogue Annotation with Speaker Characteristics Leveraging a Frozen LLM 提出通过冻结LLM增强对话注释以解决说话者特征识别问题 large language model foundation model
12 COPO: Causal-Oriented Policy Optimization for Hallucinations of MLLMs 提出COPO以解决多模态大语言模型的幻觉问题 large language model multimodal
13 RealTalk-CN: A Realistic Chinese Speech-Text Dialogue Benchmark With Cross-Modal Interaction Analysis 提出RealTalk-CN以解决中文对话系统缺乏真实语音数据的问题 large language model multimodal
14 Unveiling Over-Memorization in Finetuning LLMs for Reasoning Tasks 揭示大规模语言模型微调中的过度记忆现象 large language model instruction following
15 Transferring Expert Cognitive Models to Social Robots via Agentic Concept Bottleneck Models 提出基于概念瓶颈模型的社交机器人以增强会议效果 foundation model multimodal
16 ZARA: Zero-shot Motion Time-Series Analysis via Knowledge and Retrieval Driven LLM Agents 提出ZARA框架以解决零-shot人类活动识别问题 large language model ✅
17 GeRe: Towards Efficient Anti-Forgetting in Continual Learning of LLM via General Samples Replay 提出GeRe框架以高效解决大语言模型的遗忘问题 large language model ✅
18 EMSEdit: Efficient Multi-Step Meta-Learning-based Model Editing 提出EMSEdit以解决低数据环境下模型编辑效率问题 large language model ✅
19 Factor Augmented Supervised Learning with Text Embeddings 提出AEALT框架以解决高维文本嵌入的效率问题 large language model
20 RCR-Router: Efficient Role-Aware Context Routing for Multi-Agent LLM Systems with Structured Memory 提出RCR-Router以解决多智能体LLM系统中的上下文路由问题 large language model
21 Persistent Instability in LLM's Personality Measurements: Effects of Scale, Reasoning, and Conversation History 提出PERSIST框架以评估LLM个性测量的不稳定性 large language model
22 P-Aligner: Enabling Pre-Alignment of Language Models via Principled Instruction Synthesis 提出P-Aligner以实现语言模型的预对齐指令生成 large language model
23 TURA: Tool-Augmented Unified Retrieval Agent for AI Search 提出TURA以解决动态信息检索的实时性问题 large language model
24 Modelling and Classifying the Components of a Literature Review 提出新注释方案以提升文献综述生成质量 large language model
25 HarmonyGuard: Toward Safety and Utility in Web Agents via Adaptive Policy Enhancement and Dual-Objective Optimization 提出HarmonyGuard以解决网络代理的安全与效用平衡问题 large language model ✅
26 I Think, Therefore I Am Under-Qualified? A Benchmark for Evaluating Linguistic Shibboleth Detection in LLM Hiring Evaluations 提出语言标记检测基准以评估LLM招聘评估中的偏见问题 large language model
27 Balancing Stylization and Truth via Disentangled Representation Steering 提出StyliTruth以解决风格化与真实性之间的权衡问题 large language model
28 Semantic Bridge: Universal Multi-Hop Question Generation via AMR-Driven Graph Synthesis 提出语义桥接框架以解决多跳推理问题 large language model
29 KVSink: Understanding and Enhancing the Preservation of Attention Sinks in KV Cache Quantization for LLMs 提出KVSink以增强KV缓存量化中注意力汇聚的保护 large language model
30 Guided Navigation in Knowledge-Dense Environments: Structured Semantic Exploration with Guidance Graphs 提出Guidance Graph以解决知识密集环境中的导航问题 large language model
31 ConvMix: A Mixed-Criteria Data Augmentation Framework for Conversational Dense Retrieval 提出ConvMix框架以解决对话密集检索的数据稀缺问题 large language model
32 Confidence-Weighted Token Set Cover for Early Hypothesis Pruning in Self-Consistency 提出基于置信度加权的假设修剪方法以提高自一致性效率 chain-of-thought

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
33 IFDECORATOR: Wrapping Instruction Following Reinforcement Learning with Verifiable Rewards 提出IFDecorator以解决RLVR训练效率低下问题 reinforcement learning large language model instruction following
34 Difficulty-Based Preference Data Selection by DPO Implicit Reward Gap 提出基于难度的数据选择策略以提升偏好数据效率 reinforcement learning RLHF DPO
35 GTPO and GRPO-S: Token and Sequence-Level Reward Shaping with Policy Entropy 提出动态熵加权机制以解决长链推理中的奖励分配问题 reinforcement learning reward shaping large language model
36 Sculptor: Empowering LLMs with Cognitive Agency via Active Context Management 提出Sculptor以解决长上下文处理中的干扰问题 reinforcement learning large language model instruction following
37 Are Today's LLMs Ready to Explain Well-Being Concepts? 提出基于LLM的评估框架以提升幸福感概念解释质量 DPO direct preference optimization large language model
38 Share Your Attention: Transformer Weight Sharing via Matrix-based Dictionary Learning 提出MASA框架以解决变换器层间冗余问题 distillation large language model
39 Graph Representation Learning with Massive Unlabeled Data for Rumor Detection 利用大规模未标记数据提升谣言检测的图表示学习 representation learning
40 ShoppingBench: A Real-World Intent-Grounded Shopping Benchmark for LLM-based Agents 提出ShoppingBench以解决复杂购物意图评估问题 reinforcement learning distillation

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
41 Eliciting and Analyzing Emergent Misalignment in State-of-the-Art Large Language Models 提出MISALIGNMENTBENCH以解决大型语言模型的对齐脆弱性问题 manipulation large language model

⬅️ 返回 cs.CL 首页 · 🏠 返回主页