Research

Paper

AI LLM February 26, 2026

A Model-Free Universal AI

Authors

Yegon Kim, Juho Lee

Abstract

In general reinforcement learning, all established optimal agents, including AIXI, are model-based, explicitly maintaining and using environment models. This paper introduces Universal AI with Q-Induction (AIQI), the first model-free agent proven to be asymptotically $\varepsilon$-optimal in general RL. AIQI performs universal induction over distributional action-value functions, instead of policies or environments like previous works. Under a grain of truth condition, we prove that AIQI is strong asymptotically $\varepsilon$-optimal and asymptotically $\varepsilon$-Bayes-optimal. Our results significantly expand the diversity of known universal agents.

Metadata

arXiv ID: 2602.23242
Provider: ARXIV
Primary Category: cs.AI
Published: 2026-02-26
Fetched: 2026-02-27 04:35

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2602.23242v1</id>\n    <title>A Model-Free Universal AI</title>\n    <updated>2026-02-26T17:21:16Z</updated>\n    <link href='https://arxiv.org/abs/2602.23242v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2602.23242v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>In general reinforcement learning, all established optimal agents, including AIXI, are model-based, explicitly maintaining and using environment models. This paper introduces Universal AI with Q-Induction (AIQI), the first model-free agent proven to be asymptotically $\\varepsilon$-optimal in general RL. AIQI performs universal induction over distributional action-value functions, instead of policies or environments like previous works. Under a grain of truth condition, we prove that AIQI is strong asymptotically $\\varepsilon$-optimal and asymptotically $\\varepsilon$-Bayes-optimal. Our results significantly expand the diversity of known universal agents.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-02-26T17:21:16Z</published>\n    <arxiv:primary_category term='cs.AI'/>\n    <author>\n      <name>Yegon Kim</name>\n    </author>\n    <author>\n      <name>Juho Lee</name>\n    </author>\n  </entry>"
}