Disentangling Memory and Reasoning Ability in Large Language Models

TL;DR

This paper introduces a new inference paradigm for large language models that separates memory recall from reasoning, improving interpretability and performance by explicitly guiding the model through distinct knowledge retrieval and logical reasoning steps.

Contribution

It proposes a novel inference method that decomposes LLM reasoning into memory recall and reasoning, using special tokens to improve transparency and accuracy.

Findings

Enhanced model interpretability through explicit step separation

Improved reasoning accuracy and reliability

Facilitated error analysis and response refinement

Abstract

Large Language Models (LLMs) have demonstrated strong performance in handling complex tasks requiring both extensive knowledge and reasoning abilities. However, the existing LLM inference pipeline operates as an opaque process without explicit separation between knowledge retrieval and reasoning steps, making the model's decision-making process unclear and disorganized. This ambiguity can lead to issues such as hallucinations and knowledge forgetting, which significantly impact the reliability of LLMs in high-stakes domains. In this paper, we propose a new inference paradigm that decomposes the complex inference process into two distinct and clear actions: (1) memory recall: which retrieves relevant knowledge, and (2) reasoning: which performs logical steps based on the recalled knowledge. To facilitate this decomposition, we introduce two special tokens memory and reason, guiding the model to distinguish between steps that require knowledge retrieval and those that involve reasoning. Our experiment results show that this decomposition not only improves model performance but also enhances the interpretability of the inference process, enabling users to identify sources of error and refine model responses effectively. The code is available at https://github.com/MingyuJ666/Disentangling-Memory-and-Reasoning.