Cogito, ergo sum: A Neurobiologically-Inspired Cognition-Memory-Growth System for Code Generation

ACL ARR 2025 May Submission6507 Authors

Abstract:

Large language model-based Multi-Agent Systems (MAS) have demonstrated promising performance for enhancing the efficiency and accuracy of code generation tasks. However, most existing methods follow a conventional sequence of planning, coding, and debugging, which contradicts the growth-driven nature of the human learning process. Additionally, the frequent information interaction between multiple agents inevitably involves high computational costs.

In this paper, we propose \textbf{Cogito}, a neurobiologically-inspired multi-agent framework to enhance problem-solving capabilities in code generation tasks with lower cost. Specifically, \textbf{Cogito} adopts a reverse sequence: it first undergoes debugging, then coding, and finally planning. This approach mimics human learning and development, where knowledge is acquired progressively. Accordingly, a hippocampus-like memory module with different functions is designed to work with the pipeline to provide quick retrieval in similar tasks.

Through this growth-based learning model, \textbf{Cogito} accumulates knowledge and cognitive skills at each stage, ultimately forming a Super-Role---an all-capable agent to perform the code generation task. Extensive experiments against representative baselines demonstrate the superior performance and efficiency of \textbf{Cogito}. The code is publicly available at \url{https://anonymous.4open.science/r/test_80EF}.

Paper Type: Long

Research Area: Language Modeling

Research Area Keywords: chain-of-thought, code models, LLM/AI agents, prompting

Languages Studied: English

Submission Number: 6507