爱可可 AI 前沿推介(5.20)

文章预览

LG - 机器学习 CV - 计算机视觉 CL - 计算与语言 AS - 音频与语音 RO - 机器人 1、[CL] Is the Pope Catholic? Yes，the Pope is Catholic. Generative Evaluation of Intent Resolution in LLMs 2、[CL] LMD3：Language Model Data Density Dependence 3、[LG] Kolmogorov-Arnold Networks (KANs) for Time Series Analysis 4、[CV] No Time to Waste：Squeeze Time into Channel for Mobile Video Understanding 5、[CV] LayGA：Layered Gaussian Avatars for Animatable Clothing Transfer 摘要：LLM意图解析的生成式评价、语言模型数据密度依赖、Kolmogorov-Arnold网络(KAN)用于时间序列分析、面向移动视频理解将时间压缩到通道维度、面向可动画着装迁移的分层高斯化身 1、[CL] Is the Pope Catholic? Yes, the Pope is Catholic. Generative Evaluation of Intent Resolution in LLMs A Yerukola, S Vaduguru, D Fried, M Sap [CMU] LLM意图解析的生成式评价 要点: 提出一种新的生成式评估框架，用于评估LLM通过对非字面 ………………………………