Logic Distillation: Learning from Code Function by Function for Decision-making Tasks

Large language models (LLMs) have garnered increasing attention owing to their powerful comprehension and generation capabilities. Generally, larger LLMs (L-LLMs) that require paid interfaces exhibit significantly superior performance compared to smaller LLMs (S-LLMs) that can be deployed on a variety of devices. Knowledge distillation (KD) aims to empower S-LLMs with the capabilities of L-LLMs, while S-LLMs merely mimic the outputs of L-LLMs, failing to get the powerful decision-making capability for new situations. Consequently, S-LLMs are helpless when it comes to continuous decision-making tasks that require logical reasoning. To tackle the identified challenges, we propose a novel framework called Logic Distillation (LD). Initially, LD employs L-LLMs to instantiate complex instructions into discrete functions and illustrates their usage to establish a function base. Subsequently, LD fine-tunes S-LLMs based on the function base to learn the logic employed by L-LLMs in decision-making. During testing, S-LLMs will yield decision-making outcomes, function by function, based on current states. Experiments demonstrate that with the assistance of LD, S-LLMs can achieve outstanding results in continuous decision-making tasks, comparable to, or even surpassing, those of L-LLMs. The code and data for the proposed method are provided for research purposes https: //github. com/Anfeather/Logic-Distillation.

Authors

• Dong Chen Zhejiang University
• Shilin Zhang
• Fei Gao
• Yueting Zhuang Zhejiang University
• Siliang Tang Zhejiang University
• Qidong Liu City University of Hong Kong Xi’an Jiaotong University
• Mingliang Xu School of Computer Science and Artificial Intelligence, Zhengzhou University, Zhengzhou, 450001, China; Engineering Research Center of Intelligent Swarm Systems, Ministry of Education, Zhengzhou, 450001, China; National Supercomputing Center in Zhengzhou, Zhengzhou, 450001, China;

Keywords

• Knowledge Representation and Reasoning: KRR: Learning and reasoning
• Multidisciplinary Topics and Applications: MTA: Game playing
• Natural Language Processing: NLP: Language models