Audit-Certified Conformal Routing for Reliable Hybrid Training

Hybrid post-training combines scalar-reward updates with distillation, but noisy reward estimates can route examples into wrong-direction updates. We study a split-conformal sign gate with an independent audit of wrong-sign risk among routed examples. A single global conformal radius induces exactly a magnitude threshold; its contribution is a finite-sample marginal interpretation of that threshold, not a different sample ranking. Our evaluation separates calibration, audit, routing diagnosis, policy fitting, and final testing. On GSM8K, the global gate obtains $50.0\pm 0.5$ accuracy, while studentized routing obtains $50.7\pm 0.5$, versus $49.8\pm 0.5$ for an equally audited magnitude gate; sample-adaptive conformal variants are also evaluated on MATH, HumanEval, and two additional 3B backbones. A separate group-aware construction addresses the sign of the centered advantage used by GRPO. These guarantees concern routing on the stated distribution, not optimization gains or robustness to arbitrary policy drift.