Dense Is Not Enough: Hierarchical Supervision Allocation for Long-Horizon On-Policy Distillation
On-policy distillation (OPD) transfers the capabilities of a large language model to a smaller student by providing teacher supervision on the student's own rollouts. In long-horizon agentic tasks, however, uniform token-level matching can allocate supervision poorly: a large local discrepancy need not improve future b...
Yu-Hao Sun, Bin-Rui Wu, Zhuo-Er Xu et al.
· 0 citations