TY  - RPRT
TI  - When Every Simulation Counts: Value-Based Reinforcement Learning for Accelerated Photonics Inverse Design
AU  - Longying Wen
AU  - Feiyang Wu
AU  - Jinglin Yu
AU  - Chongxian Yuan
AU  - Renjie Li
AU  - Zhaoyu Zhang
PY  - 2026
UR  - https://arxiv.org/abs/2607.23469
ID  - 2607.23469
ER  -