ProCreations/repro-estimating-continuous-treatment-effects-with-two-stage-kernel-ridge-regression

Repair Sobolev lower bound with finite hard prior

Expand KRR scope and rerun Job Corps validation

Claim 1: run the estimator instead of auditing the formula; separate the treatment-effect and conditional-response complexities and show the rate tracks only the former

claims 2-3: build the two-stage KRR estimator and measure convergence rates + CV adaptation

Synchronize Hub LFS rule and publish validated ICML reproduction for ziqS4yXFQX

Publish validated ICML reproduction for ziqS4yXFQX

initial commit