cs.CL（2025-07-03）

📊 共 21 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) (13) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗1) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (13 篇)

#题目一句话要点标签🔗⭐
1 Coling-UniA at SciVQA 2025: Few-Shot Example Retrieval and Confidence-Informed Ensembling for Multimodal Large Language Models 针对科学视觉问答，提出基于少样本检索和置信度加权集成的多模态大语言模型方案。 large language model multimodal
2 SynapseRoute: An Auto-Route Switching Framework on Dual-State Large Language Model SynapseRoute：双状态大语言模型上的自动路由切换框架，优化医疗问答。 large language model
3 Large Language Models for Automating Clinical Data Standardization: HL7 FHIR Use Case 利用大型语言模型GPT-4o和Llama 3.2实现临床数据向HL7 FHIR格式的半自动化转换。 large language model
4 ReliableMath: Benchmark of Reliable Mathematical Reasoning on Large Language Models ReliableMath：评估大型语言模型在数学推理中可靠性的基准 large language model
5 Enhancing Temporal Sensitivity of Large Language Model for Recommendation with Counterfactual Tuning 提出CETRec，通过因果推理增强LLM在推荐中对时序信息的敏感性 large language model
6 DeepGesture: A conversational gesture synthesis system based on emotions and semantics DeepGesture：基于情感和语义的会话手势合成系统 large language model multimodal
7 Is Reasoning All You Need? Probing Bias in the Age of Reasoning Language Models 推理语言模型更易受社会偏见影响：CLEAR-Bias基准测试揭示安全性隐患 large language model chain-of-thought
8 DeSTA2.5-Audio: Toward General-Purpose Large Audio Language Model with Self-Generated Cross-Modal Alignment DeSTA2.5-Audio：通过自生成跨模态对齐实现通用大型音频语言模型 large language model instruction following
9 Revisiting Active Learning under (Human) Label Variation 提出一种考虑人类标注差异的主动学习框架，提升真实场景下的标注效率。 large language model
10 MPF: Aligning and Debiasing Language Models post Deployment via Multi Perspective Fusion MPF：通过多视角融合实现部署后语言模型的对齐和去偏 large language model
11 From Measurement to Mitigation: Exploring the Transferability of Debiasing Approaches to Gender Bias in Maltese Language Models 探索偏见消除方法在马耳他语语言模型中的迁移性，以缓解性别偏见 large language model
12 Dynamic Long Short-Term Memory Based Memory Storage For Long Horizon LLM Interaction 提出Pref-LSTM，利用BERT分类器和LSTM记忆模块，为长程LLM交互实现动态记忆存储。 large language model
13 Efficient Code LLM Training via Distribution-Consistent and Diversity-Aware Data Selection 提出基于分布一致性和多样性感知的数据选择方法，提升代码大语言模型训练效率。 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🔗⭐
14 Multimodal Mathematical Reasoning with Diverse Solving Perspective 提出MathV-DP数据集与Qwen-VL-DP模型，提升多模态数学推理能力 reinforcement learning large language model multimodal
15 Self-Correction Bench: Uncovering and Addressing the Self-Correction Blind Spot in Large Language Models 揭示并解决大语言模型自我纠错盲点，提升安全关键应用可靠性 reinforcement learning large language model
16 Generalizing Verifiable Instruction Following 提出IFBench基准以解决指令跟随泛化问题 reinforcement learning instruction following
17 RLVER: Reinforcement Learning with Verifiable Emotion Rewards for Empathetic Agents 提出RLVER框架，利用可验证情感奖励提升LLM的共情能力 reinforcement learning PPO large language model
18 ARF-RLHF: Adaptive Reward-Following for RLHF through Emotion-Driven Self-Supervision and Trace-Biased Dynamic Optimization ARF-RLHF：通过情感驱动的自监督和轨迹偏置动态优化，实现自适应奖励跟随 PPO RLHF DPO
19 MOTIF: Modular Thinking via Reinforcement Fine-tuning in LLMs 提出MOTIF，通过强化学习微调LLM，实现模块化思维以突破上下文长度限制。 reinforcement learning large language model ✅
20 Rewrite-to-Rank: Optimizing Ad Visibility via Retrieval-Aware Text Rewriting 提出Rewrite-to-Rank框架，通过重写广告文本优化其在检索系统中的可见性。 reinforcement learning PPO

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
21 Adversarial Manipulation of Reasoning Models using Internal Representations 利用内部表征对抗操纵推理模型，发现并利用“谨慎”方向进行越狱攻击 manipulation chain-of-thought ✅

⬅️ 返回 cs.CL 首页 · 🏠 返回主页