Beyond Words: A Latent Memory Approach to Internal Reasoning in LLMs

TL;DR

This paper introduces an implicit memory framework for LLMs that enhances internal reasoning efficiency by mimicking human mental representations, showing significant training loss reduction and maintaining interpretability.

Contribution

It proposes a novel implicit memory module for LLMs, integrating human-like mental representations to improve reasoning efficiency and robustness.

Findings

Implicit Memory Module reduces training loss by 35-57%.

Framework maintains interpretability with optional chain-of-thought decoding.

Discusses scalable mechanisms and theoretical foundations for internal reasoning.

Abstract

Recent advances in large language models (LLMs) have popularized the chain-of-thought (CoT) paradigm, in which models produce explicit reasoning steps in natural language. Although this approach improves interpretability and facilitates external auditing, it may not represent the most computationally efficient method for internal reasoning. In contrast, human cognition relies on implicit mental representations that recall past sensory and episodic information without requiring complete verbalization. In this paper, we propose a framework that integrates implicit mental representations into the internal reasoning processes of LLMs. Preliminary experiments indicate that incorporating an Implicit Memory Module (IMM) into a simple GPT model yields a reduction of between 35% and 57% in final training loss compared to a regular GPT baseline. The addition of an explicit interpretability channel (e.g., a chain-of-thought decoder) is straightforward to implement within this approach. We outline theoretical foundations, propose technical mechanisms to scale the memory module, and discuss how these ideas may lead to more efficient and robust reasoning, with optional future extensions for explicit auditability.