CoTAR: Chain-of-Thought Attribution Reasoning with Multi-level Granularity
Moshe Berchansky, Daniel Fleischer, Moshe Wasserblat, Peter Izsak · 2024
State-of-the-art performance in QA tasks is currently achieved by systems employing Large Language Models (LLMs), however these models tend to hallucinate information in their responses.One approach focuses on enhancing the generation process by incorporating attribution from the given input to the output.However, the challenge of identifying appropriate attributions and verifying their accuracy against a source is a complex task that requires significant improvements in assessing such systems.We introduce an attribution-oriented Chain-of-Thought reasoning method to enhance the accuracy of attributions.This approach focuses the reasoning process on generating an attributioncentric output.Evaluations on two contextenhanced question-answering datasets using GPT-4 demonstrate improved accuracy and correctness of attributions.In addition, the combination of our method with finetuning enhances the response and attribution accuracy of two smaller LLMs, showing their potential to outperform GPT-4 in some cases. 1 1 Our code is publicly available for reproduction: https: //github.com/mosheber/cotar.git what is the title of tears for fears song?[1] Johnny Panic and the Bible of Dreams...[3] ... Songs from the Big Chair: ...[2] Human eyes produce tears... QuestionLets analyze the relevant spans: From passage [1]: * Johnny Panic and the ... * is a song by the British band ... From passage [3]: * international hit singles * "Mothers Talk", "Shout", .... Thus, the final answer is: