Conference
Open access
2026
Don't Be Misled by Style: A Style-Adaptive Reranker for Capturing Effective Knowledge in Retrieval-Augmented Generation
SARK is proposed, a style-augmented multi-task framework that prioritizes effective knowledge over stylistic perturbations in the reranker model and improves generation performance across multiple LLMs under mixed-style conditions.
Ruwen Zhang, Bo Liu, Zhang-Sheng Xiang et al.
· Annual Meeting of the Associ... · 0 citations