Songchun Zhang Zeyue Xue Siming Fu Jie Huang 0017 Xianghao Kong Y. Ma Haoyang Huang Nan Duan 0001 Anyi Rao Astrolabe: Steering Forward-Process Reinforcement Learning for Distilled Autoregressive Video Models. 2026 March abs/2603.17051 CoRR https://doi.org/10.48550/arXiv.2603.17051 db/journals/corr/corr2603.html#abs-2603-17051 streams/journals/corr