What a Data Release Must Preserve for Off-Policy Evaluation

Abstract:In off-policy evaluation, merging decision-time histories that matter differently for a specified policy value or comparison can make that evaluation unidentified, even with observed actions and outcomes and known behavior and target policies. We characterize which distinctions must be preserved: identification holds if and only if the action-specific evaluation weights are constant within each released record, and the coarsest exact release groups histories by these weights. For normalized outcomes, the largest sharp width induced by the release equals a behavior-weighted within-record radius of these weights. The same radius becomes a release-design objective when each record may be refined into at most $m$ categories. The optimum has an exact $m$-center characterization. When behavior does not vary within a released record, we give a finite-data repair algorithm whose maximum evaluation-weight radius is at most twice the finite-data optimum. For a fixed action-outcome distribution within a released record, a distinct sharp-identification problem remains. Under the same condition, the action-specific outcome distributions are identified, and optimizing over their couplings gives the sharp interval. We construct an AIPW envelope containing this sharp interval; it matches the interval for all such outcome distributions exactly when the extreme within-record action-coefficient vectors take at most two values in each action coordinate. In 24 KuaiRec tasks, the repair algorithm yields a 93% median reduction in mean pairwise sharp width at $m=64$ relative to the unrepaired release.

Machine Learning (stat.ML); Machine Learning (cs.LG)

arXiv:2610.15283 [stat.ML]

(or arXiv:2610.15283v1 [stat.ML] for this version)

https://doi.org/10.48550/arXiv.2610.15283 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

expected/aipw_rows.csv

expected/figure1_summary.csv

expected/figure2_summary.csv

expected/heartsteps_phase.csv

kuairec_task_manifest.csv

requirements.txt

src/heartsteps.py

src/kuairec/aipw_summary.py

src/kuairec/aipw_task.py

src/kuairec/common.py

src/kuairec/design_summary.py

src/kuairec/design_task.py

src/kuairec/evaluation_summary.py

src/kuairec/evaluation_task.py

src/kuairec/matched_information.py

src/kuairec/mcenter.py

src/kuairec/repair_common.py

src/kuairec/repair_evaluation_summary.py

src/kuairec/repair_evaluation_task.py

src/kuairec/repair_summary.py

src/kuairec/repair_task.py

src/kuairec/sharp.py

src/reproduce.py

(21 additional files not shown) You must enabled JavaScript to view entire file list.

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.