cs.CL（2025-07-05）

📊 共 12 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) (9) 支柱二：RL算法与架构 (RL & Architecture) (2 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (9 篇)

#题目一句话要点标签🔗⭐
1 Large Language Models for Zero-Shot Multicultural Name Recognition 提出Prompt-Engineered Fine-Tuning框架，解决零样本跨文化姓名识别难题。 large language model
2 Losing our Tail -- Again: On (Un)Natural Selection And Multilingual Large Language Models 多语言大模型导致语言多样性衰退：警惕（非）自然选择与模型坍塌 large language model
3 An HTR-LLM Workflow for High-Accuracy Transcription and Analysis of Abbreviated Latin Court Hand 提出四阶段工作流程以提高中世纪法律文件的转录精度 large language model multimodal
4 Detecting Token-Level Hallucinations Using Variance Signals: A Reference-Free Approach 提出一种基于方差信号的无参考Token级幻觉检测方法 large language model
5 OrthoRank: Token Selection via Sink Token Orthogonality for Efficient LLM inference OrthoRank：利用Sink Token正交性的Token选择方法，提升LLM推理效率。 large language model
6 SymbolicThought: Integrating Language Models and Symbolic Reasoning for Consistent and Interpretable Human Relationship Understanding SymbolicThought：结合语言模型与符号推理，实现一致且可解释的人际关系理解 large language model
7 Nunchi-Bench: Benchmarking Language Models on Cultural Reasoning with a Focus on Korean Superstition Nunchi-Bench：提出针对韩国迷信文化的LLM文化推理能力评测基准。 large language model
8 Demystifying ChatGPT: How It Masters Genre Recognition 研究揭示ChatGPT在电影类型识别任务中超越其他LLM的卓越能力 large language model
9 Do LLMs Overthink Basic Math Reasoning? Benchmarking the Accuracy-Efficiency Tradeoff in Language Models 提出Overthinking Score，评估LLM在基础数学推理中的准确率-效率权衡 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)

#题目一句话要点标签🔗⭐
10 Conversation Forests: The Key to Fine Tuning Large Language Models for Multi-Turn Medical Conversations is Branching 提出Savage Conversation Forests，用于微调LLM以提升多轮医疗对话诊断准确率。 reinforcement learning DPO direct preference optimization
11 Handling Korean Out-of-Vocabulary Words with Phoneme Representation Learning 提出KOPL框架，利用音素表征学习处理韩语未登录词问题 representation learning ✅

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

#题目一句话要点标签🔗⭐
12 XISM: an eXploratory and Interactive Graph Tool to Visualize and Evaluate Semantic Map Models XISM：一种交互式图工具，用于可视化和评估语义地图模型 semantic map ✅

⬅️ 返回 cs.CL 首页 · 🏠 返回主页