Retroformer: Retrospective Large Language Agents with Policy Gradient Optimization
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2023
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910434372616192 |
|---|---|
| author | Yao, Weiran Heinecke, Shelby Niebles, Juan Carlos Liu, Zhiwei Feng, Yihao Xue, Le Murthy, Rithesh Chen, Zeyuan Zhang, Jianguo Arpit, Devansh Xu, Ran Mui, Phil Wang, Huan Xiong, Caiming Savarese, Silvio |
| author_facet | Yao, Weiran Heinecke, Shelby Niebles, Juan Carlos Liu, Zhiwei Feng, Yihao Xue, Le Murthy, Rithesh Chen, Zeyuan Zhang, Jianguo Arpit, Devansh Xu, Ran Mui, Phil Wang, Huan Xiong, Caiming Savarese, Silvio |
| contents | Recent months have seen the emergence of a powerful new trend in which large language models (LLMs) are augmented to become autonomous language agents capable of performing objective oriented multi-step tasks on their own, rather than merely responding to queries from human users. Most existing language agents, however, are not optimized using environment-specific rewards. Although some agents enable iterative refinement through verbal feedback, they do not reason and plan in ways that are compatible with gradient-based learning from rewards. This paper introduces a principled framework for reinforcing large language agents by learning a retrospective model, which automatically tunes the language agent prompts from environment feedback through policy gradient. Specifically, our proposed agent architecture learns from rewards across multiple environments and tasks, for fine-tuning a pre-trained language model which refines the language agent prompt by summarizing the root cause of prior failed attempts and proposing action plans. Experimental results on various tasks demonstrate that the language agents improve over time and that our approach considerably outperforms baselines that do not properly leverage gradients from the environment. This demonstrates that using policy gradient optimization to improve language agents, for which we believe our work is one of the first, seems promising and can be applied to optimize other models in the agent architecture to enhance agent performances over time. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2308_02151 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Retroformer: Retrospective Large Language Agents with Policy Gradient Optimization Yao, Weiran Heinecke, Shelby Niebles, Juan Carlos Liu, Zhiwei Feng, Yihao Xue, Le Murthy, Rithesh Chen, Zeyuan Zhang, Jianguo Arpit, Devansh Xu, Ran Mui, Phil Wang, Huan Xiong, Caiming Savarese, Silvio Computation and Language Artificial Intelligence Recent months have seen the emergence of a powerful new trend in which large language models (LLMs) are augmented to become autonomous language agents capable of performing objective oriented multi-step tasks on their own, rather than merely responding to queries from human users. Most existing language agents, however, are not optimized using environment-specific rewards. Although some agents enable iterative refinement through verbal feedback, they do not reason and plan in ways that are compatible with gradient-based learning from rewards. This paper introduces a principled framework for reinforcing large language agents by learning a retrospective model, which automatically tunes the language agent prompts from environment feedback through policy gradient. Specifically, our proposed agent architecture learns from rewards across multiple environments and tasks, for fine-tuning a pre-trained language model which refines the language agent prompt by summarizing the root cause of prior failed attempts and proposing action plans. Experimental results on various tasks demonstrate that the language agents improve over time and that our approach considerably outperforms baselines that do not properly leverage gradients from the environment. This demonstrates that using policy gradient optimization to improve language agents, for which we believe our work is one of the first, seems promising and can be applied to optimize other models in the agent architecture to enhance agent performances over time. |
| title | Retroformer: Retrospective Large Language Agents with Policy Gradient Optimization |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2308.02151 |