FinePrompt: Unveiling the Role of Finetuned Inductive Bias on Compositional Reasoning in GPT-4
Submission Type: Regular Short Paper
Submission Track: Theme Track: Large Language Models and the Future of NLP
Submission Track 2: Question Answering
Keywords: Large Language Model, Prompt Learning, Fine-tuning, Compositional Reasoning, Question Answering
TL;DR: This study focuses on enhancing GPT-4's compositional reasoning abilities by transferring task-specific inductive biases from fine-tuned models to prompts
Abstract: Compositional reasoning across texts has been a long-standing challenge in natural language processing. With large language models like GPT-4 taking over the field, prompting techniques such as chain-of-thought (CoT) were proposed to unlock compositional, multi-step reasoning capabilities of LLMs. Despite their success, the prompts demand significant human effort to discover and validate them. Our work draws attention to the idea of transferring task-specific inductive biases from finetuned models to prompts, as a way of improving GPT-4's compositional reasoning capabilities. To leverage these inductive biases, we formulate prompt templates to ease the transfer of inductive biases. The experimental results on multi-hop question answering and numerical reasoning over text show that our proposed prompt scheme shows competitive zero-shot and few-shot performances compared to existing prompts on complicated reasoning tasks, highlighting the importance of adopting the validated biases of the previous paradigm.
Submission Number: 1902
Loading