ReGuLaR: Variational Latent Reasoning Guided by Rendered Chain-of-Thought

TL;DR

ReGuLaR introduces a novel variational latent reasoning approach guided by rendered reasoning chains, significantly improving efficiency and effectiveness of reasoning in large language models, and surpassing traditional Chain-of-Thought methods.

Contribution

The paper presents a new latent learning paradigm using rendered reasoning chains as visual inputs to guide variational auto-encoding, enhancing compression and reasoning performance.

Findings

ReGuLaR outperforms existing latent reasoning methods in efficiency and accuracy.

ReGuLaR surpasses Chain-of-Thought in multi-modal reasoning tasks.

The approach effectively compresses reasoning processes with minimal information loss.

Abstract

While Chain-of-Thought (CoT) significantly enhances the performance of Large Language Models (LLMs), explicit reasoning chains introduce substantial computational redundancy. Recent latent reasoning methods attempt to mitigate this by compressing reasoning processes into latent space, but often suffer from severe performance degradation due to the lack of appropriate compression guidance. In this study, we propose Rendered CoT-Guided variational Latent Reasoning (ReGuLaR), a simple yet novel latent learning paradigm resolving this issue. Fundamentally, we formulate latent reasoning within the Variational Auto-Encoding (VAE) framework, sampling the current latent reasoning state from the posterior distribution conditioned on previous ones. Specifically, when learning this variational latent reasoning model, we render explicit reasoning chains as images, from which we extract dense visual-semantic representations to regularize the posterior distribution, thereby achieving efficient compression with minimal information loss. Extensive experiments demonstrate that ReGuLaR significantly outperforms existing latent reasoning methods across both computational efficiency and reasoning effectiveness, and even surpasses CoT through multi-modal reasoning, providing a new and insightful solution to latent reasoning. Code: https://github.com/FanmengWang/ReGuLaR.