Towards A Unified View of Answer Calibration for Multi-Step Reasoning
This work provides a unified framework for researchers and practitioners to optimize multi-step reasoning in AI, though it is incremental as it synthesizes existing methods rather than introducing new ones.
The paper tackles the lack of systematic analysis of answer calibration techniques in multi-step reasoning with LLMs, categorizing them into step-level and path-level strategies and finding that combining both yields optimal results.
Large Language Models (LLMs) employing Chain-of-Thought (CoT) prompting have broadened the scope for improving multi-step reasoning capabilities. We generally divide multi-step reasoning into two phases: path generation to generate the reasoning path(s); and answer calibration post-processing the reasoning path(s) to obtain a final answer. However, the existing literature lacks systematic analysis on different answer calibration approaches. In this paper, we summarize the taxonomy of recent answer calibration techniques and break them down into step-level and path-level strategies. We then conduct a thorough evaluation on these strategies from a unified view, systematically scrutinizing step-level and path-level answer calibration across multiple paths. Experimental results reveal that integrating the dominance of both strategies tends to derive optimal outcomes. Our study holds the potential to illuminate key insights for optimizing multi-step reasoning with answer calibration.