Research

Paper

AI LLM March 02, 2026

Recursive Models for Long-Horizon Reasoning

Authors

Chenxiao Yang, Nathan Srebro, Zhiyuan Li

Abstract

Modern language models reason within bounded context, an inherent constraint that poses a fundamental barrier to long-horizon reasoning. We identify recursion as a core principle for overcoming this barrier, and propose recursive models as a minimal realization, where the model can recursively invoke itself to solve subtasks in isolated contexts. We prove that any computable problem admits a recursive decomposition in which each subtask requires only exponentially smaller active context than standard autoregressive models; this strictly surpasses any context management approach confined to a single sequence, such as summarization. We further generalize our framework to modern agentic systems with arbitrary context processing and control flows, and prove that recursive models can achieve optimal power within this broader class. Experimentally, we train a 3B model to reason recursively and evaluate on Boolean satisfiability, a task requiring long-horizon combinatorial search, where it significantly outperforms frontier LLMs.

Metadata

arXiv ID: 2603.02112
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-03-02
Fetched: 2026-03-03 04:34

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.02112v1</id>\n    <title>Recursive Models for Long-Horizon Reasoning</title>\n    <updated>2026-03-02T17:37:10Z</updated>\n    <link href='https://arxiv.org/abs/2603.02112v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.02112v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Modern language models reason within bounded context, an inherent constraint that poses a fundamental barrier to long-horizon reasoning. We identify recursion as a core principle for overcoming this barrier, and propose recursive models as a minimal realization, where the model can recursively invoke itself to solve subtasks in isolated contexts. We prove that any computable problem admits a recursive decomposition in which each subtask requires only exponentially smaller active context than standard autoregressive models; this strictly surpasses any context management approach confined to a single sequence, such as summarization. We further generalize our framework to modern agentic systems with arbitrary context processing and control flows, and prove that recursive models can achieve optimal power within this broader class. Experimentally, we train a 3B model to reason recursively and evaluate on Boolean satisfiability, a task requiring long-horizon combinatorial search, where it significantly outperforms frontier LLMs.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n    <published>2026-03-02T17:37:10Z</published>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Chenxiao Yang</name>\n    </author>\n    <author>\n      <name>Nathan Srebro</name>\n    </author>\n    <author>\n      <name>Zhiyuan Li</name>\n    </author>\n  </entry>"
}