Skip to content
Sagan

Paper

Training-Inference Kernel Contracts: Bounding Divergence in Post-Training and Deployment

Unreadunread

AI summary

arXiv:2606. 07581v1 Announce Type: new Abstract: A modern post-training pipeline often writes one symbol for its policy, pi_theta, while evaluating it through two different programs: a training kernel optimized for autograd and an inference kernel optimized for low-precision, fused, dynamically batched serving.