First-Step Advantage: Importance of Starting Right in Multi-Step Math Reasoning
This addresses the issue of error propagation in reasoning for smaller language models, though it is incremental as it builds on existing chain-of-thought methods.
The paper tackles the problem of language models making errors in multi-step math reasoning due to incorrect initial steps, and shows that guiding smaller models to start correctly leads to significant performance gains, such as up to +24 points on GSM8K for 7B models.
Language models can solve complex reasoning tasks better by learning to generate rationales for their predictions. Often these models know how to solve a task but their auto-regressive decoding nature leads to incorrect results if they start incorrectly. We observe that smaller models in particular when corrected, can solve a task that they would have otherwise struggled with. We demonstrate this phenomenon by using a larger model to guide smaller models, which leads to significantly improved performance (up to +24 points on the GSM8K dataset by 7B models). To assist smaller models in initiating the starting step, we propose QuestCoT, where a smaller model first asks itself how to start, before proceeding with a chain of reasoning. On various multistep mathematical reasoning datasets over multiple smaller models, we show that getting the right start can lead to significant performance gains across all models (gains of up to +6 points on GSM8K, +9 on SVAMP, +5 on ASDiv, and +7 on MultiArith).