English

LLM2: Let Large Language Models Harness System 2 Reasoning

Computation and Language 2025-03-03 v2 Artificial Intelligence

Abstract

Large language models (LLMs) have exhibited impressive capabilities across a myriad of tasks, yet they occasionally yield undesirable outputs. We posit that these limitations are rooted in the foundational autoregressive architecture of LLMs, which inherently lacks mechanisms for differentiating between desirable and undesirable results. Drawing inspiration from the dual-process theory of human cognition, we introduce LLM2, a novel framework that combines an LLM (System 1) with a process-based verifier (System 2). Within LLM2, the LLM is responsible for generating plausible candidates, while the verifier provides timely process-based feedback to distinguish desirable and undesirable outputs. The verifier is trained with a pairwise comparison loss on synthetic process-supervision data generated through our token quality exploration strategy. Empirical results on mathematical reasoning benchmarks substantiate the efficacy of LLM2, exemplified by an accuracy enhancement from 50.3 to 57.8 (+7.5) for Llama3-1B on GSM8K. Furthermore, when combined with self-consistency, LLM2 achieves additional improvements, boosting major@20 accuracy from 56.2 to 70.2 (+14.0).

Keywords

Cite

@article{arxiv.2412.20372,
  title  = {LLM2: Let Large Language Models Harness System 2 Reasoning},
  author = {Cheng Yang and Chufan Shi and Siheng Li and Bo Shui and Yujiu Yang and Wai Lam},
  journal= {arXiv preprint arXiv:2412.20372},
  year   = {2025}
}

Comments

Accepted to NAACL 2025 Main Conference

R2 v1 2026-06-28T20:50:58.972Z