cs.CL（2025-07-13）

📊 共 6 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

#	题目	一句话要点	标签	🔗
1	Large Language Models Encode Semantics in Low-Dimensional Linear Subspaces	发现大语言模型在低维线性子空间中编码语义信息，并提出基于此的防御机制。	large language model
2	Towards Agentic RAG with Deep Reasoning: A Survey of RAG-Reasoning Systems in LLMs	综述Agentic RAG：深度推理增强LLM的检索增强生成系统	large language model multimodal	✅
3	How Important is `Perfect' English for Machine Translation Prompts?	研究用户提示中的错误对LLM机器翻译性能的影响	large language model instruction following
4	MCEval: A Dynamic Framework for Fair Multilingual Cultural Evaluation of LLMs	MCEval：用于评估LLM多语言文化公平性的动态框架	large language model

#	题目	一句话要点	标签	🔗	⭐
5	ViSP: A PPO-Driven Framework for Sarcasm Generation with Contrastive Learning	提出ViSP框架，结合PPO和对比学习，用于生成高质量多模态讽刺文本。	PPO contrastive learning large language model	✅
6	Your Pretrained Model Tells the Difficulty Itself: A Self-Adaptive Curriculum Learning Paradigm for Natural Language Understanding	提出基于预训练模型自适应难度评估的课程学习方法，提升NLU任务性能。	curriculum learning