Guaranteeing Knowledge Integration with Joint Decoding for Retrieval-Augmented Generation
Zhengyi Zhao, Shubo Zhang, Zezhong Wang, Yuxi Zhang, Huimin Wang, Yutian Zhao, Yefeng Zheng, Binyang Li, Kam-Fai Wong, Xian Wu · 2026
Read Guaranteeing Knowledge Integration with Joint Decoding for Retrieval-Augmented Generation by Zhengyi Zhao, Shubo Zhang, Zezhong Wang, Yuxi Zhang, Huimin Wang, Yutian Zhao, Yefeng Zheng, Binyang Li, Kam-Fai Wong, Xian Wu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-08046.txt.
Description
Zhao and nine colleagues tackle the integration bottleneck in RAG, where retrieved files often fail to change model answers, even when matching. Current research stresses lookup quality but overlooks a key failure: models ignore matching files that conflict with internal parametric memory. Resolving that conflict quietly in a single generation pass proves weak. Their GuarantRAG framework openly splits reasoning from evidence integration. An Inner-Answer from built-in knowledge alone captures the model's reasoning flow, while a Refer-Answer built with a contrastive training goal ensures faithful evidence use.
Catalog record
| CREATOR | Zhengyi Zhao, Shubo Zhang, Zezhong Wang, Yuxi Zhang, Huimin Wang, Yutian Zhao, Yefeng Zheng, Binyang Li, Kam-Fai Wong, Xian Wu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 87106 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.08046