CoRE: Enhancing Metacognition with Label-free Self-evaluation in LRMs

TL;DR

This paper introduces CoRE, a label-free self-evaluation method for large reasoning models that improves reasoning efficiency and accuracy by detecting cyclical reasoning patterns without external labels.

Contribution

The paper proposes CoRE, a novel latent space embedding for self-evaluation in LRMs, and CoRE-Eval, a training-free framework to detect redundant reasoning, enhancing efficiency and accuracy.

Findings

Reduces reasoning steps by up to 33.2%

Improves accuracy by around 10% on benchmarks

Achieves 70% accuracy on AIME with 32B model

Abstract

Large reasoning models (LRMs) have demonstrated impressive capabilities in domains like mathematics and program synthesis. Despite their strong performance, LRMs often exhibit overthinking -- excessive and redundant reasoning steps that introduce inefficiencies during inference. This phenomenon raises an important question for LRM self-evaluation: How can a model autonomously assess the correctness of its own reasoning trajectory without external labels? To address this, we propose Chain-of-Reasoning Embedding (CoRE), a series of hidden states in latent space to enable label-free self-evaluation on intermediate reasoning steps of LRMs, so as to enhance metacognition abilities for improved reasoning efficiency. By analyzing the geometric properties of the CoRE trajectories, we reveal that redundant reasoning usually presents cyclical fluctuations, which correspond to repetitive and unconscious reflection/exploration. Leveraging this insight, we further introduce a training-free, label-free self-evaluation framework, CoRE-Eval, to detect such patterns and dynamically determine whether to terminate reasoning early. Extensive experiments on mathematical reasoning benchmarks (GSM8K, MATH-500, and AIME) and across model sizes from 7B to 32B demonstrate that CoRE-Eval reduces chain-of-thought length by 13.7% to 33.2% while improving answer accuracy by around 10%, achieving 70.0% accuracy on the challenging AIME benchmark with the 32B model.