Feedback Adaptation for Retrieval-Augmented Generation

TL;DR

This paper introduces feedback adaptation as a new evaluation framework for RAG systems, emphasizing how quickly and effectively they incorporate corrective feedback during deployment.

Contribution

It proposes two metrics for measuring feedback adaptation, and introduces PatchRAG, a method for immediate correction without retraining.

Findings

Training-based approaches show a trade-off between correction delay and adaptation reliability.

PatchRAG achieves immediate correction and strong generalization after feedback.

Feedback adaptation is a crucial but overlooked aspect of RAG system behavior.

Abstract

Retrieval-Augmented Generation (RAG) systems are typically evaluated under static assumptions, despite being frequently corrected through user or expert feedback in deployment. Existing evaluation protocols focus on overall accuracy and fail to capture how systems adapt after feedback is introduced. We introduce feedback adaptation as a problem setting for RAG systems, which asks how effectively and how quickly corrective feedback propagates to future queries. To make this behavior measurable, we propose two evaluation axes: correction lag, which captures the delay between feedback provision and behavioral change, and post-feedback performance, which measures reliability on semantically related queries after feedback. Using these metrics, we show that training-based approaches exhibit a trade-off between delayed correction and reliable adaptation. We further propose PatchRAG, a minimal inference-time instantiation that incorporates feedback without retraining, demonstrating immediate correction and strong post-feedback generalization under the proposed evaluation. Our results highlight feedback adaptation as a previously overlooked dimension of RAG system behavior in interactive settings.