cs.CL（2023-12-09）

📊 共 8 篇论文

🎯 兴趣领域导航

#	题目	一句话要点	标签
1	Beneath the Surface: Unveiling Harmful Memes with Multimodal Reasoning Distilled from Large Language Models	提出基于大语言模型蒸馏的多模态推理框架，用于识别有害Meme	large language model multimodal
2	Aligner: One Global Token is Worth Millions of Parameters When Aligning Large Language Models	Aligner：利用单个全局Token对齐大语言模型，参数效率提升显著	large language model instruction following
3	Two Directions for Clinical Data Generation with Large Language Models: Data-to-Label and Label-to-Data	提出基于大语言模型的临床数据生成方法以检测阿尔茨海默病相关症状	large language model
4	History Matters: Temporal Knowledge Editing in Large Language Model	提出METO框架，解决大语言模型中时序知识编辑的历史知识遗忘问题	large language model
5	Steering Llama 2 via Contrastive Activation Addition	提出对比激活添加(CAA)方法，通过操纵激活向量来引导Llama 2的行为。	large language model

#	题目	一句话要点	标签
6	Understanding the Effect of Model Compression on Social Bias in Large Language Models	研究模型压缩对大型语言模型社会偏见的影响，发现量化具有正则化效果。	distillation large language model
7	Enhancing Robustness of Foundation Model Representations under Provenance-related Distribution Shifts	通过溯源调整增强基础模型表征在分布偏移下的鲁棒性	predictive model foundation model
8	PILLOW: Enhancing Efficient Instruction Fine-tuning via Prompt Matching	PILLOW：通过提示匹配增强高效指令微调	reinforcement learning large language model