TY  - RPRT
TI  - DUET: Optimize Token-Budget Allocation for Reinforcement Learning with Verifiable Rewards
AU  - Haoyu Hu
AU  - Xuandong Zhao
AU  - Xuhai "Orson'' Xu
AU  - Nori Jacoby
PY  - 2026
UR  - https://arxiv.org/abs/2605.08441
ID  - 2605.08441
ER  -