The paper introduces Adaptive Doubly Robust (ADR), an off‑policy evaluation method for ranking policies that blends adaptive importance weighting with reward regression to reduce variance. ADR is unbiased when the true user behavior model is known and, under a sufficient condition, achieves lower variance than the prior Adaptive Inverse Propensity Scoring (AIPS) approach. Experiments on synthetic data show that ADR consistently improves mean squared error over AIPS and other ranking OPE estimators across various data sizes and ranking lengths.
By Kosuke Iguchi, Ren Kishimoto
arXiv:2610.08677v1 Announce Type: new
Abstract: Off-policy evaluation (OPE) for contextual bandit policies becomes challenging when action-level importance weighting incurs excessive variance. Doubly...
By Nicol\`o Felicioni, Michael Benigni, Maurizio Ferrari Dacrema, Paolo Cremonesi
Off-policy evaluation (OPE) for contextual bandit policies becomes challenging when action-level importance weighting incurs excessive variance. Doubly robust (DR) estimation remains unbiased under co...
arXiv:2606. 19117v1 Announce Type: cross Abstract: Offline policy learning has received growing attention in causal inference.
By Yiyan Huang, Cheuk Hang Leung, Qi Wu, Zhiheng Zhang
The paper introduces a doubly robust framework for distributionally robust offline contextual pricing, addressing challenges of distributional shifts and continuous actions. It proposes a localized DR estimator that is computationally efficient and semiparametrically optimal, and a learning algorithm that exploits demand noise smoothness to achieve finite‑sample regret bounds of τ^{-σ/(2σ+1)} for smoothness orders s=1,2. Experiments across varying distribution shifts demonstrate the effectiveness of the proposed methods.
By Min Xu, Xinyi Yin, Yunfan Zhang, Yuxuan Han, Houcai Shen, Caihua Chen
The paper introduces Prediction-Powered Ranking (PP‑Rank), a framework that blends large language model (LLM) judge predictions with limited human feedback to rank LLMs more precisely while maintaining valid inference under adaptive sampling. PP‑Rank uses data‑driven prediction weights to minimize asymptotic covariance, establishes asymptotic normality for general evaluation targets, and provides a rule for reducing exploration that balances cost and variance. Experiments on UltraFeedback‑Extended and synthetic data show that PP‑Rank improves score estimation, narrows confidence intervals, certifies rankings, and can cut evaluation costs without losing precision.
By Gabriel Sargent, Will Wei Sun, Zhengwu Zhang, Yufeng Liu