# Recursive Language Models

> Recursive Language Models (RLMs) extend language model capabilities by enabling recursive, compositional reasoning over complex hierarchical structures.

Source: https://metavert.io/recursive-language-models  
Updated: 2026-03-10

**Recursive Language Models (RLMs)** are an emerging architecture that extends [large language models](https://metavert.io/large-language-models) with the ability to reason recursively—breaking complex problems into subproblems, solving each independently, and composing the results. Where standard autoregressive models generate tokens left-to-right in a flat sequence, RLMs can decompose and recompose at multiple levels of abstraction.

The core insight is that many reasoning tasks are inherently hierarchical. Writing a complex document involves structuring sections, then paragraphs, then sentences. Solving a math problem requires decomposing it into sub-problems. Planning a multi-step task means nesting goals within goals. Standard [transformers](https://metavert.io/transformer-architecture) handle this implicitly through [attention](https://metavert.io/attention-mechanism) patterns, but RLMs make the recursion explicit—allowing the model to "call itself" on sub-tasks and integrate the results.

This approach addresses a fundamental limitation of current LLMs: the difficulty of consistent, deep compositional reasoning. A model generating a 10,000-word document in a single forward pass must hold the entire structure in its [context window](https://metavert.io/context-windows) simultaneously. An RLM can instead plan the high-level structure, then recursively expand each section, maintaining coherence through hierarchical composition rather than brute-force attention.

RLMs connect to broader themes in AI architecture. They echo the structure of recursive programs in computer science, the hierarchical planning used in [AI agent](https://metavert.io/agentic-ai) frameworks, and the compositional nature of human cognition. As [reasoning models](https://metavert.io/reasoning-models) push toward more complex autonomous tasks, recursive architectures may prove essential for scaling beyond the current limits of flat autoregressive generation.

## Related Topics

- [Large Language Models](https://metavert.io/large-language-models)
- [Reasoning Models](https://metavert.io/reasoning-models) [Test-Time Compute](https://metavert.io/test-time-compute)
- [Transformer Architecture](https://metavert.io/transformer-architecture)
- [Attention Mechanism](https://metavert.io/attention-mechanism)
- [Context Windows](https://metavert.io/context-windows)
- [AI Agents](https://metavert.io/agentic-ai)

## Further Reading

- [Recursive Language Models (arXiv)](https://arxiv.org/abs/2512.24601)