PhD student, MPI-SWS
1 paper at NeurIPS 2025
We propose a curriculum strategy for guiding the training of agents that operate under strict trajectory constraints during deployment by adaptively tightening constraints based on agent's performance.