Improving Self Consistency in LLMs through Probabilistic Tokenization
This addresses the underutilization of multiple tokenization capabilities in modern LLMs to enhance reasoning consistency, representing an incremental improvement.
The paper tackles the problem of improving self-consistency in large language models (LLMs) for reasoning tasks by leveraging probabilistic tokenization, which involves using multiple tokenizations of the same input during training. The result shows that this approach generates logically diverse reasoning paths, with experiments conducted on 5 LLM families and 4 reasoning benchmarks.
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.