Beyond Meta-Reasoning: Metacognitive Consolidation for Self-Improving LLM Reasoning

TL;DR

This paper introduces Metacognitive Consolidation, a framework enabling LLMs to accumulate and reuse meta-reasoning experience over time, leading to improved reasoning performance across benchmarks.

Contribution

It proposes a novel hierarchical framework for consolidating meta-reasoning experience, moving beyond episodic approaches to enable continuous self-improvement in LLM reasoning.

Findings

Performance gains observed across multiple benchmarks.

Meta-knowledge accumulation improves reasoning over time.

Rich, attributable meta-level traces enhance meta-reasoning.

Abstract

Large language models (LLMs) have demonstrated strong reasoning capabilities, and as existing approaches for enhancing LLM reasoning continue to mature, increasing attention has shifted toward meta-reasoning as a promising direction for further improvement. However, most existing meta-reasoning methods remain episodic: they focus on executing complex meta-reasoning routines within individual instances, but ignore the accumulation of reusable meta-reasoning skills across instances, leading to recurring failure modes and repeatedly high metacognitive effort. In this paper, we introduce Metacognitive Consolidation, a novel framework in which a model consolidates metacognitive experience from past reasoning episodes into reusable knowledge that improves future meta-reasoning. We instantiate this framework by structuring instance-level problem solving into distinct roles for reasoning, monitoring, and control to generate rich, attributable meta-level traces. These traces are then consolidated through a hierarchical, multi-timescale update mechanism that gradually forms evolving meta-knowledge. Experimental results demonstrate consistent performance gains across benchmarks and backbone models, and show that performance improves as metacognitive experience accumulates over time.