PageLLM: A Multi-Grained Reward Framework for Whole-Page Optimization with Large Language Models 文章

ArXiv CS.AI2026-05-26NEWSen作者: Xinyuan Wang, Liang Wu, Dongjie Wang, Yanjie Fu
大语言模型
查看原文 →
关系图谱

PageLLM: A Multi-Grained Reward Framework for Whole-Page Optimization with Large Language Models · 相关技术

相关技术

LLMODEPLAlanguage model递归自我改进alignmentTURNSULReferring expression comprehension (REC)Parts-of-Speech (POS) tagsOTALMMlarge language modelsHISForFFIEffort Metric AttentionDAPTARGANNICE