cs.CL（2026-08-19）

📊 共 10 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

#	题目	一句话要点	标签	🔗
1	MedUAG: Unified Understanding and Generation for Medical Multimodal Models	提出MedUAG以解决医疗多模态模型的理解与生成问题	large language model multimodal
2	Gradient Mirage: Trainable yet Label-Unidentifiable Gradients in Large Language Model Split Learning	提出Gradient Mirage以解决大语言模型分割学习中的梯度匹配攻击问题	large language model
3	Do Large Language Models Hallucinate Electric Fata Morganas?	探讨大型语言模型的幻觉现象及其哲学意义	large language model
4	Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs	研究量化对大语言模型主动干扰的影响	large language model	✅
5	Identifying Implicit Premises for Logical Reconstruction of Argument Graphs	提出神经符号管道以识别隐含前提解决论证图重构问题	large language model
6	Execution-grounded evaluation reveals hidden failures in language-model calculations for environmental science	提出AtmosCoder-Bench以揭示环境科学语言模型计算中的隐性错误	large language model
7	Beyond LLM-Based Reasoning: Lightweight GNNs for Agent Failure Attribution	提出AFANet以解决多智能体系统中的故障归因问题	large language model
8	Shared Circuits for Shared Grammar: Tracing Subject-Verb Agreement Across Languages	提出共享电路以追踪跨语言主谓一致性问题	large language model

#	题目	一句话要点	标签	🔗	⭐
9	Multimodal Rapport Estimation in Real-World HRI	提出多模态方法以解决真实场景中的人机互动质量评估问题	JEPA multimodal
10	DART-SD: Diamond-topology Aware Retrieval and Tuning for Self-Distillation of Multi-Turn Tool-Calling Agents	提出DART-SD以解决多回合工具调用代理的轨迹模仿问题	distillation large language model