A Hopfieldian View-based Interpretation for Chain-of-Thought Reasoning
This addresses the need for rigorous explanations of CoT's success in AI reasoning, which is incremental as it builds on existing CoT methods with a novel interpretative framework.
The paper tackles the problem of explaining why Chain-of-Thought (CoT) reasoning improves performance in large language models, proposing a Hopfieldian view-based interpretation and a Read-and-Control approach that demonstrates the ability to decipher CoT's inner workings, provide error localization, and control reasoning paths across seven datasets for three tasks.
Chain-of-Thought (CoT) holds a significant place in augmenting the reasoning performance for large language models (LLMs). While some studies focus on improving CoT accuracy through methods like retrieval enhancement, yet a rigorous explanation for why CoT achieves such success remains unclear. In this paper, we analyze CoT methods under two different settings by asking the following questions: (1) For zero-shot CoT, why does prompting the model with "let's think step by step" significantly impact its outputs? (2) For few-shot CoT, why does providing examples before questioning the model could substantially improve its reasoning ability? To answer these questions, we conduct a top-down explainable analysis from the Hopfieldian view and propose a Read-and-Control approach for controlling the accuracy of CoT. Through extensive experiments on seven datasets for three different tasks, we demonstrate that our framework can decipher the inner workings of CoT, provide reasoning error localization, and control to come up with the correct reasoning path.