爱可可 AI 前沿推介(11.15)

文章预览

LG - 机器学习 CV - 计算机视觉 CL - 计算与语言 RO - 机器人 1、[CL] Dynamic Subset Tuning：Expanding the Operational Range of Parameter-Efficient Training for Large Language Models 2、[CL] Dynamic Rewarding with Prompt Optimization Enables Tuning-free Self-Alignment of Language Models 3、[IR] A Large-Scale Study of Relevance Assessments with Large Language Models：An Initial Look 4、[RO] Offline Adaptation of Quadruped Locomotion using Diffusion Models 5、[LG] LLMStinger：Jailbreaking LLMs using RL fine-tuned LLMs 摘要：扩展大型语言模型参数高效训练的操作范围、提示优化动态奖励使语言模型的自对齐无需调优、基于大型语言模型的相关性评估的大规模研究、基于扩散模型的四足机器人运动离线自适应、用强化学习微调LLM来越狱LLM 1、[CL] Dynamic Subset Tuning: Expanding the Operational Range of Parameter-Efficient Training for Large Language Models F Stahlberg, J Lichtarge, S Kumar [Googl ………………………………