Improving Self Consistency in LLMs through Probabilistic Tokenization

TL;DR

This paper introduces a novel approach to improve the self-consistency of large language models in reasoning tasks by leveraging probabilistic tokenization capabilities, which have been underutilized in modern LLM training.

Contribution

The paper proposes a new method to utilize probabilistic tokenizations in LLMs, enhancing their reasoning consistency and generating more logically diverse reasoning paths.

Findings

Probabilistic tokenization improves LLM self-consistency.

Enhanced reasoning diversity beyond surface-level linguistic variation.

Consistent improvements across 5 LLMs and 4 benchmarks.

Abstract

Prior research has demonstrated noticeable performance gains through the use of probabilistic tokenizations, an approach that involves employing multiple tokenizations of the same input string during the training phase of a language model. Despite these promising findings, modern large language models (LLMs) have yet to be trained using probabilistic tokenizations. Interestingly, while the tokenizers of these contemporary LLMs have the capability to generate multiple tokenizations, this property remains underutilized. In this work, we propose a novel method to leverage the multiple tokenization capabilities of modern LLM tokenizers, aiming to enhance the self-consistency of LLMs in reasoning tasks. Our experiments indicate that when utilizing probabilistic tokenizations, LLMs generate logically diverse reasoning paths, moving beyond mere surface-level linguistic diversity.We carefully study probabilistic tokenization and offer insights to explain the self consistency improvements it brings through extensive experimentation on 5 LLM families and 4 reasoning benchmarks.