大模型时代下可信递归自优化基础、风险、评估与开放性挑战.pptxVIP

  • 0
  • 0
  • 约27.97万字
  • 约 10页
  • 2026-11-04 发布于河北
  • 举报

大模型时代下可信递归自优化基础、风险、评估与开放性挑战.pptx

TrustworthyRecursiveSelf-Improvement

intheEraofLargeModels

Foundation,Risks,EvaluationandOpenChallenges

JiayuePu1,AngGao1,JinyiBai1,QipengWang1,TongyuWen1,JiahaoZhao1,KeyuQu1,JunhaoZhang2,Chenglei

Shen1,†,ZhongxiangSun1,†,JunXu1andXiaoZhang1,*

1GaolingSchoolofArtificialIntelligence,RenminUniversityofChina,2UniversityofChineseAcademyofSciences,†Projectsupervisor,

*Correspondingauthor

AIsystemsbuiltaroundlargemodelsincreasinglyuseexperienceandfeedbacktomakepersistentchangestomodels,prompts,memories,tools,andworkflows.

文档评论(0)

1亿VIP精品文档

相关文档