From Search to Sampling: Generative Models for Robust Algorithmic Recourse

2025-05-12

International Conference on Learning Representations

AI Summary

Existing counterfactual recourse methods optimize proximity, plausibility, and validity separately during training, leading to suboptimal joint search during inference. This paper proposes GenRe, the first framework to jointly model all three objectives end-to-end. GenRe builds a generative counterfactual recommender based on a conditional variational autoencoder; introduces an unsupervised supervision signal synthesis mechanism, with theoretical proof of its consistency as an estimator; and replaces fragile gradient-based optimization with forward sampling for robust and efficient inference. Evaluated on multiple benchmarks, GenRe achieves state-of-the-art performance in the holistic trade-off among recourse cost, plausibility, and validity. The implementation is publicly available.

Technology Category

• Natural Language Processing: Generation
• Reasoning under Uncertainty: Causality
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

Algorithmic Recourse provides recommendations to individuals who are adversely impacted by automated model decisions, on how to alter their profiles to achieve a favorable outcome. Effective recourse methods must balance three conflicting goals: proximity to the original profile to minimize cost, plausibility for realistic recourse, and validity to ensure the desired outcome. We show that existing methods train for these objectives separately and then search for recourse through a joint optimization over the recourse goals during inference, leading to poor recourse recommendations. We introduce GenRe, a generative recourse model designed to train the three recourse objectives jointly. Training such generative models is non-trivial due to lack of direct recourse supervision. We propose efficient ways to synthesize such supervision and further show that GenRe's training leads to a consistent estimator. Unlike most prior methods, that employ non-robust gradient descent based search during inference, GenRe simply performs a forward sampling over the generative model to produce minimum cost recourse, leading to superior performance across multiple metrics. We also demonstrate GenRe provides the best trade-off between cost, plausibility and validity, compared to state-of-art baselines. Our code is available at: https://github.com/prateekgargx/genre.

Problem

• Balancing proximity, plausibility, and validity in algorithmic recourse
• Overcoming poor recourse from separate training of objectives
• Generating robust recourse via joint training and forward sampling

Innovation

• Generative model jointly trains three recourse objectives
• Synthesizes supervision for training without direct recourse data
• Forward sampling replaces gradient descent for robust recourse