Calibrated for Which Policy? Conformal Uncertainty for Fixed World Models Under Intervention

arXiv:2610.35877v1 Announce Type: new Abstract: Changing a deployment policy can invalidate a world model's uncertainty while leaving the physical transition mechanism and predictor unchanged. For one-step prediction, if source and target laws share the conditional mechanism P(Y|X)=Q(Y|X), then the intervention is identified by the pre-outcome change of measure w(X)=dQ_X/dP_X. For policy-controlled dynamics this ratio factors into state-occupancy and action-propensity terms. Weighted conformal calibration transports marginal coverage to the target law; with an estimated ratio hat w, P_Q{Y in C_{hat w}(X)} ge 1-alpha-TV(Q,hat Q) ge 1-alpha-E_P|w-hat w|. Population log-loss regret for a source-versus-target domain classifier bounds the same ratio error. If h is its clipped target posterior, rho is the target-domain prior, and Delta(h) is population log-loss regret, then E_P|w-hat w_h| le (1/(rhogamma))sqrt{Delta(h)/2}. An exact cancellation improves the dependence from a generic odds-map bound of order gamma^{-2} to (rhogamma)^{-1}. Our experiments use a released DINO-feature world model and physical Franka trajectories. Under the strongest controlled state-action shift, ordinary 90% conformal sets cover 55.0%, whereas oracle and learned-ratio calibration cover 91.0% and 90.9%. The learned surrogate remains within 0.9% total variation of the target law. Calibration is relative to the deployment law: the model can remain fixed while calibration follows the intervention. We machine-check in Lean 4 the finite-support algebraic and causal identities used in the classifier-regret argument, including its coverage-composition step.

Source: arXiv cs.LG | 2026-10-28