Reinforcement Retrieval Leveraging Fine-grained Feedback for Fact Checking News Claims with Black-Box LLM

TL;DR

This paper introduces FFRR, a reinforcement learning approach that improves news claim fact-checking by leveraging fine-grained feedback from black-box LLMs to optimize retrieval strategies.

Contribution

It presents a novel two-level feedback mechanism that enhances retrieval for fact-checking using reinforcement learning with black-box LLMs.

Findings

FFRR outperforms baseline models on news claim verification datasets.

Fine-grained feedback improves retrieval accuracy.

Reinforcement learning effectively optimizes retrieval policies.

Abstract

Retrieval-augmented language models have exhibited promising performance across various areas of natural language processing (NLP), including fact-critical tasks. However, due to the black-box nature of advanced large language models (LLMs) and the non-retrieval-oriented supervision signal of specific tasks, the training of retrieval model faces significant challenges under the setting of black-box LLM. We propose an approach leveraging Fine-grained Feedback with Reinforcement Retrieval (FFRR) to enhance fact-checking on news claims by using black-box LLM. FFRR adopts a two-level strategy to gather fine-grained feedback from the LLM, which serves as a reward for optimizing the retrieval policy, by rating the retrieved documents based on the non-retrieval ground truth of the task. We evaluate our model on two public datasets for real-world news claim verification, and the results demonstrate that FFRR achieves significant improvements over strong LLM-enabled and non-LLM baselines.