Chain-of-Action: Faithful and Multimodal Question Answering through Large Language Models
This addresses the issue of unreliable and limited reasoning in QA systems for users needing accurate, real-time information from diverse sources, representing a novel method for a known bottleneck.
The paper tackles the problems of unfaithful hallucination and weak reasoning in multimodal and retrieval-augmented question answering by introducing the Chain-of-Action (CoA) framework, which uses a novel reasoning-retrieval mechanism and achieves improved performance over other methods on public benchmarks and a Web3 case study.
We present a Chain-of-Action (CoA) framework for multimodal and retrieval-augmented Question-Answering (QA). Compared to the literature, CoA overcomes two major challenges of current QA applications: (i) unfaithful hallucination that is inconsistent with real-time or domain facts and (ii) weak reasoning performance over compositional information. Our key contribution is a novel reasoning-retrieval mechanism that decomposes a complex question into a reasoning chain via systematic prompting and pre-designed actions. Methodologically, we propose three types of domain-adaptable `Plug-and-Play' actions for retrieving real-time information from heterogeneous sources. We also propose a multi-reference faith score (MRFS) to verify and resolve conflicts in the answers. Empirically, we exploit both public benchmarks and a Web3 case study to demonstrate the capability of CoA over other methods.