The authors frame multilingual query rewriting as an open-ended planning problem rather than a one-size-fits-all translation step.
Adaptive query rewriting narrows language gaps in multilingual reasoning
A reinforcement learning agent learns to select rewriting strategies per query, outperforming static translation-based baselines.
Chinese Tech
Rui Qi · Yufeng Chen · Yunlong Liang · Chuan Meng · Sijin Lu · Ge Shi · +3 more
Beijing Jiaotong University · Tencent · University of Edinburgh
Research Digest··2 min read
The authors introduce mRewriter-R1, an agentic framework that treats multilingual query rewriting as a multi-turn sequential decision process.
Why this paper
From Tencent and 2 others
In one line
An agentic multilingual query rewriter using reinforcement learning dynamically selects rewriting operators to improve reasoning across languages.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ·No benchmark numbers found
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§