arXiv cs.AI· Shobhita Sundaram, Cansu Sancaktar, Phillip Isola·· 23 小时前AI 评分58
LLM 智能体还需要被告知如何自我改进吗?
Do LLM Agents Still Need to Be Told How to Self-Improve?
AI 导读
论文研究一种自主范式:给定基础原语、资源和笼统的自我改进目标,让智能体自行构建基于上下文的自我学习算法。作者以游戏为测试环境,给冻结的 LLM 智能体一定算力预算和收集经验、更新文本文档的工具,不设外部循环,也不提供真实世界数据。
正文 · 原文
Abstract:LLM agents increasingly drive their own learning. Self-improvement methods delegate various components of model optimization to the models themselves. As models become more capable, this motivates a basic question: do these increasingly capable models still need us to specify how they should self-improve? In this paper we study an autonomous paradigm in which, given basic primitives, resources, and a general goal for self-improvement, agents construct their own self-learning algorithm for context-based optimization. Using the setting of gameplay as a testbed, a frozen LLM agent is given a compute budget and tools for collecting experience and updating text documents, and allowed to freely experiment without an externally imposed loop or access to real-world data. We find that sufficiently capable models do self-improve under these conditions, demonstrating strategies that align with classic cognitive frameworks for self-regulated learning and constructs from machine learning literature on self-generated curricula. Furthermore, we study the extent to which the learned text artifacts transfer to external benchmarks, observing improvement across different game horizons, mechanics, and 3D navigation.
| Comments: | Project page: this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.23219 [cs.LG] |
| (or arXiv:2611.23219v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2611.23219 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Shobhita Sundaram [view email]
[v1]
Tue, 17 Nov 2026 18:57:10 UTC (1,807 KB)
来源:arXiv cs.AI · arxiv.org