TY  - RPRT
TI  - Open Materials Generation with Inference-Time Reinforcement Learning
AU  - Philipp Hoellmer
AU  - Stefano Martiniani
PY  - 2026
UR  - https://arxiv.org/abs/2602.00424
ID  - 2602.00424
ER  -