Hugging Face Papers ↗ · 2026-11-17
大模型智能体还需要被告知如何自我改进吗?
Media reports Hugging Face Papers Source language: Chinese
论文摘要(作者报告,未经本站复测):大模型智能体越来越多地驱动自身的学习。自我改进方法将模型优化的各个组件委托给模型自身。
Full text
Capable frozen LLM agents can autonomously construct context-based self-learning algorithms from basic resources and a general goal, yielding text artifacts that transfer to new game and 3D navigation tasks.