Lessons Learned: A Multi-Agent Framework for Code LLMs to Learn and Improve
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909865507553280 |
|---|---|
| author | Liu, Yuanzhe Deng, Ryan Kaler, Tim Chen, Xuhao Leiserson, Charles E. Ma, Yao Chen, Jie |
| author_facet | Liu, Yuanzhe Deng, Ryan Kaler, Tim Chen, Xuhao Leiserson, Charles E. Ma, Yao Chen, Jie |
| contents | Recent studies show that LLMs possess different skills and specialize in different tasks. In fact, we observe that their varied performance occur in several levels of granularity. For example, in the code optimization task, code LLMs excel at different optimization categories and no one dominates others. This observation prompts the question of how one leverages multiple LLM agents to solve a coding problem without knowing their complementary strengths a priori. We argue that a team of agents can learn from each other's successes and failures so as to improve their own performance. Thus, a lesson is the knowledge produced by an agent and passed on to other agents in the collective solution process. We propose a lesson-based collaboration framework, design the lesson solicitation--banking--selection mechanism, and demonstrate that a team of small LLMs with lessons learned can outperform a much larger LLM and other multi-LLM collaboration methods. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_23946 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Lessons Learned: A Multi-Agent Framework for Code LLMs to Learn and Improve Liu, Yuanzhe Deng, Ryan Kaler, Tim Chen, Xuhao Leiserson, Charles E. Ma, Yao Chen, Jie Artificial Intelligence Machine Learning Multiagent Systems Software Engineering Recent studies show that LLMs possess different skills and specialize in different tasks. In fact, we observe that their varied performance occur in several levels of granularity. For example, in the code optimization task, code LLMs excel at different optimization categories and no one dominates others. This observation prompts the question of how one leverages multiple LLM agents to solve a coding problem without knowing their complementary strengths a priori. We argue that a team of agents can learn from each other's successes and failures so as to improve their own performance. Thus, a lesson is the knowledge produced by an agent and passed on to other agents in the collective solution process. We propose a lesson-based collaboration framework, design the lesson solicitation--banking--selection mechanism, and demonstrate that a team of small LLMs with lessons learned can outperform a much larger LLM and other multi-LLM collaboration methods. |
| title | Lessons Learned: A Multi-Agent Framework for Code LLMs to Learn and Improve |
| topic | Artificial Intelligence Machine Learning Multiagent Systems Software Engineering |
| url | https://arxiv.org/abs/2505.23946 |