General Purpose Verification for Chain of Thought Prompting
This addresses the challenge of unreliable reasoning in LLMs for tasks requiring step-by-step logic, though it is incremental as it builds on existing chain-of-thought methods.
The paper tackled the problem of improving reasoning capabilities in Large Language Models by exploring different chains of thought and validating individual steps for relevance, mathematical accuracy, and logical consistency, resulting in better accuracy than vanilla generation and outperforming best-of-N sampling in 6 out of 9 datasets.
Many of the recent capabilities demonstrated by Large Language Models (LLMs) arise primarily from their ability to exploit contextual information. In this paper, we explore ways to improve reasoning capabilities of LLMs through (1) exploration of different chains of thought and (2) validation of the individual steps of the reasoning process. We propose three general principles that a model should adhere to while reasoning: (i) Relevance, (ii) Mathematical Accuracy, and (iii) Logical Consistency. We apply these constraints to the reasoning steps generated by the LLM to improve the accuracy of the final generation. The constraints are applied in the form of verifiers: the model itself is asked to verify if the generated steps satisfy each constraint. To further steer the generations towards high-quality solutions, we use the perplexity of the reasoning steps as an additional verifier. We evaluate our method on 4 distinct types of reasoning tasks, spanning a total of 9 different datasets. Experiments show that our method is always better than vanilla generation, and, in 6 out of the 9 datasets, it is better than best-of N sampling which samples N reasoning chains and picks the lowest perplexity generation.