| Oct 07, 2026 | Released Q-Learning with Scalar Adjoint Matching (SQAM), a cheaper off-policy RL algorithm for flow policies. arXiv · blog · code |
| Sep 27, 2026 | I will join the Autonomous Learning Robots (ALR) Lab at KIT as a visiting scholar from October to December 2026, hosted by Prof. Gerhard Neumann. |
| Sep 25, 2026 | Trust Region Q Adjoint Matching (TRQAM) has been accepted to NeurIPS 2026. arXiv · blog · code |
| May 26, 2026 | Released Trust Region Q Adjoint Matching (TRQAM), a stable off-policy RL algorithm for pretrained flow policies. arXiv · blog · code |
| May 01, 2026 | Co-authored RLDX-1 Technical Report released. arXiv · code |