AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Think Then Rewrite: Reasoning Enhanced Query Rewriting for Domain Specific Retrieval

Ang Li, Yufei Shi, Yuxuan Si, Yiquan Wu, Ming Cai, Xu Tan, Yi Wang, Changlong Sun, Xiaozhong Liu, Kun Kuang

AAAI 2026 · Large language models › Retrieval and knowledge · approach: Retrieval-augmented reasoning · 1 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 CMI reward to the recall+format reward (ablation)TTR · EQUALSR@10: 43.26 -> 48.62Improved
add recall reward (ablation)TTR · EQUALSR@10: 40.98 -> 48.62Improved
add format reward (ablation)TTR · EQUALSR@10: 46.41 -> 48.62Improved
use adaptive negative sampling instead of random negatives only (ablation)TTR CMI reward · EQUALSR@10: 47.18 -> 48.62Improved
use CMI reward instead of similarity-based reward (ablation)TTR · EQUALSR@10: 46.54 -> 48.62Improved
add off-policy golden rewrites (ablation)TTR · EQUALSR@10: 39.52 -> 48.62Improved
use constructed golden rewrites instead of raw target documents as off-policy data (ablation)TTR · EQUALSR@10: 44.52 -> 48.62Improved
TTR reward and off-policy design over vanilla GRPO rewritingQwen2.5-1.5B · EQUALSR@10: 34.71 -> 48.62Improved
TTR over LUFFY (mixed-policy with recall reward only)Qwen2.5-1.5B · KUAKE-IRR@10: 42.39 -> 46.49Improved
swap GRPO for PPO/DAPO-style RL algorithmsquery rewriter · EQUALSR@10: 28.54 vs 27.05 pure retrieverNo clear effect
SFT on golden rewrites with reasoningQwen2.5-1.5B · KUAKE-IRR@10: 43.24 -> 36.04Hurt
increase CMI reward weightTTR · EQUALS and KUAKE-IRRecall@10Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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