cs.CL（2025-05-05）

📊 共 18 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) (10 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (8 🔗3)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (10 篇)

#题目一句话要点标签🔗⭐
1 Bye-bye, Bluebook? Automating Legal Procedure with Large Language Models 评估大型语言模型在自动化法律程序中的能力，揭示其在Bluebook引用格式上的局限性。 large language model
2 Rethinking Multimodal Sentiment Analysis: A High-Accuracy, Simplified Fusion Architecture 提出一种高精度、简化的多模态情感分析融合架构，在IEMOCAP数据集上达到92%的准确率。 multimodal
3 Invoke Interfaces Only When Needed: Adaptive Invocation for Large Language Models in Question Answering 提出AttenHScore，用于问答系统中大模型自适应调用，解决小模型幻觉问题。 large language model
4 LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis LLaMA-Omni2：基于LLM的实时流式语音聊天机器人 large language model instruction following
5 ReplaceMe: Network Simplification via Depth Pruning and Transformer Block Linearization 提出ReplaceMe，一种免训练的深度剪枝方法，通过线性化Transformer块简化网络。 large language model ✅
6 Memorization or Interpolation ? Detecting LLM Memorization through Input Perturbation Analysis 提出PEARL，通过输入扰动分析检测大型语言模型中的记忆化行为 large language model
7 Improving Model Alignment Through Collective Intelligence of Open-Source LLMS 提出MoAA：利用开源LLM的集体智慧提升模型对齐效果 large language model
8 UCSC at SemEval-2025 Task 3: Context, Models and Prompt Optimization for Automated Hallucination Detection in LLM Output UCSC提出基于上下文检索、模型和提示优化的框架，在LLM幻觉检测任务中取得最佳性能。 large language model
9 Predicting Movie Hits Before They Happen with LLMs 利用大语言模型预测电影冷启动期的票房表现 large language model
10 Colombian Waitresses y Jueces canadienses: Gender and Country Biases in Occupation Recommendations from LLMs 首个多语言交叉性别与国家偏见研究，揭示LLM职业推荐中的刻板印象 large language model

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

#题目一句话要点标签🔗⭐
11 RADLADS: Rapid Attention Distillation to Linear Attention Decoders at Scale RADLADS：通过快速注意力蒸馏实现线性注意力解码器的大规模应用 linear attention distillation ✅
12 RM-R1: Reward Modeling as Reasoning 提出RM-R1，通过推理增强奖励模型，提升大语言模型对齐人类偏好的能力。 reinforcement learning distillation large language model ✅
13 A Survey on Progress in LLM Alignment from the Perspective of Reward Design 综述性论文：从奖励设计的角度探讨LLM对齐的进展 reinforcement learning reward design large language model
14 Generative Sign-description Prompts with Multi-positive Contrastive Learning for Sign Language Recognition 提出GSP-MC方法，利用生成式提示和多正例对比学习提升手语识别精度。 contrastive learning large language model
15 EMORL: Ensemble Multi-Objective Reinforcement Learning for Efficient and Flexible LLM Fine-Tuning EMORL：集成多目标强化学习，高效灵活地微调大型语言模型 reinforcement learning large language model
16 Sailing by the Stars: A Survey on Reward Models and Learning Strategies for Learning from Rewards 综述：基于奖励模型的LLM奖励学习策略研究 reinforcement learning RLHF DPO ✅
17 JTCSE: Joint Tensor-Modulus Constraints and Cross-Attention for Unsupervised Contrastive Learning of Sentence Embeddings 提出JTCSE框架，通过联合张量模约束和交叉注意力提升无监督对比学习句向量表示 contrastive learning distillation
18 SIMPLEMIX: Frustratingly Simple Mixing of Off- and On-policy Data in Language Model Preference Learning SIMPLEMIX：一种简单有效的混合策略，提升语言模型偏好学习效果 preference learning DPO

⬅️ 返回 cs.CL 首页 · 🏠 返回主页