cs.CL（2023-12-16）

📊 共 7 篇论文

🎯 兴趣领域导航

#	题目	一句话要点	标签
1	One-Shot Learning as Instruction Data Prospector for Large Language Models	Nuggets：利用单样本学习从海量数据中挖掘高质量指令数据，提升大语言模型性能	large language model
2	When Parameter-efficient Tuning Meets General-purpose Vision-language Models	PETAL：面向通用视觉-语言模型的参数高效微调方法	multimodal
3	RIGHT: Retrieval-augmented Generation for Mainstream Hashtag Recommendation	提出RIGHT模型，利用检索增强生成方法解决主流话题标签推荐难题。	large language model
4	CoAScore: Chain-of-Aspects Prompting for NLG Evaluation	提出CoAScore，利用链式方面提示提升NLG评估与人类判断的相关性。	large language model
5	Continuous Prompt Generation from Linear Combination of Discrete Prompt Embeddings	提出基于离散提示嵌入的连续提示生成方法以提升可解释性	large language model

#	题目	一句话要点	标签	🔗	⭐
6	Debiasing Multimodal Sarcasm Detection with Contrastive Learning	提出对比学习框架，解决多模态讽刺检测中的文本偏见问题	contrastive learning multimodal
7	A Soft Contrastive Learning-based Prompt Model for Few-shot Sentiment Analysis	提出基于软对比学习的Prompt模型SCP，解决少样本情感分析中语义距离微妙的问题。	contrastive learning chain-of-thought