Multimodal Retrieval Augmented Generation Evaluation Benchmark
Abstract
While advances in prompt engineering and RAG have improved LLM proficiency in field-specific, specialized tasks, there has yet to be an industry standard or accepted evaluation metric of the highly fragmented RAG solutions that are currently being deployed. Thus, in this work, we focused on building a robust LLM and RAG evaluation platform. We contribute 1) a platform that evaluates an RAG system's performance on a multimodal input context for LLM question answering and 2) MRAFE: Multimodal Retrieval Augmented Feature Extractor, which processes information from the input to our platform. Through various automated and systematic hand testing, we find that our evaluation benchmarks are useful in determining noise robustness, negative rejection, information integration, and counterfactual rejection. Such a platform would serve as a useful tool for developers iterating on retrieval systems and regulatory bodies creating AI-focused governance alike.
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.