AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Faithful in Steps: Improving Generalization and Citation in RAG via Query Decomposition

Yue Liu, Zhongying Ru, Shimin Di, Jipeng Zhang, Ruiyuan Zhang, Xiaofang Zhou

AAAI 2026 · Large language models › Retrieval and knowledge · approach: RAG robustness and evaluation · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add DPO alignment with retrieval feedback to QD-R (ablation)QDRAG-R 7B · MMQAcitation (AutoAIS): 53.8 -> 57.5Improved
add DPO alignment with retrieval feedback to QD-R (ablation)QDRAG-R 7B · WebQAcitation (AutoAIS): 56.8 -> 59.9Improved
add query decomposition (D, R, S) to context reranking (ablation)QDRAG-C · MMQAcitation (AutoAIS): 48.8 -> 52.6Improved
add query decomposition (D, R, S) to context reranking (ablation)QDRAG-C · WebQAcitation (AutoAIS): 41.9 -> 45.8Improved
use query decomposition instead of query rewriting (multi-round, with DPO) (ablation)QDRAG-R · WebQAcitation (AutoAIS): 48.4 -> 59.9Improved
fine-tune on the QD task instead of the RAGQ task (ablation)QDRAG-R · InfoSeek (out-of-domain)citation (AutoAIS): 68.4 -> 72.9Improved
fine-tune on the QD task instead of the RAGQ task (ablation)QDRAG-R · E-VQA (out-of-domain)citation (AutoAIS): 26.0 -> 33.8Improved
fine-tune on RAGQ task instead of leaving the base model unfinetunedQwen2.5VL-7B RAG · InfoSeek (out-of-domain)LLM-judge: 36.7 -> 32.9Hurt
replace CLIP retriever with MMEmbed-7B retrieverQDRAG-C · MMQAcitation (AutoAIS): 52.6 -> 74.7Improved
apply sparse context filtering (reranking) rather than dense RAGQwen2.5VL-7B RAG · multi-hop MMQA and WebQAanswer accuracyHurt
reduce distractor images in the citation contextMLLM image citation · CIRCO and CIRRprecision and recallImproved
use QDRAG-C instead of SearChain for decomposition costquery decomposition · Query decomposition costcost per query (USD): 0.070 -> 0.019Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy