Moritz Hardt

Policy Design in Long-run Welfare Dynamics ICLR 2025 Limits to scalable evaluation at the frontier: LLM as judge won’t beat twice the data ICLR 2025 Training on the Test Task Confounds Evaluation and Emergence ICLR 2025 Lawma: The Power of Specialization for Legal Annotation ICLR 2025 Inherent Trade-Offs between Diversity and Stability in Multi-Task Benchmarks ICML 2024 Allocation Requires Prediction Only if Inequality Is Low ICML 2024 Causal Inference out of Control: Estimating Performativity without Treatment Randomization ICML 2024 Questioning the Survey Responses of Large Language Models NIPS 2024 An engine not a camera: Measuring performative power of online search NIPS 2024 Evaluating language models as risk scores NIPS 2024 Don’t Label Twice: Quantity Beats Quality when Comparing Binary Classifiers on a Budget ICML 2024 Causal Inference from Competing Treatments ICML 2024 Test-Time Training on Nearest Neighbors for Large Language Models ICLR 2024 Unprocessing Seven Years of Algorithmic Fairness ICLR 2024 Do causal predictors generalize better to new domains? NIPS 2024 A Theory of Dynamic Benchmarks ICLR 2023 Algorithmic Collective Action in Machine Learning ICML 2023 Performative Power NIPS 2022 Alternative Microfoundations for Strategic Classification ICML 2021 Retiring Adult: New Datasets for Fair Machine Learning NIPS 2021 Balancing Competing Objectives with Noisy Data: Score-Based Classifiers for Welfare-Aware Machine Learning ICML 2020 Strategic Classification is Causal Modeling in Disguise ICML 2020 Stochastic Optimization for Performative Prediction NIPS 2020 Linear Dynamics: Clustering without identification AISTATS 2020 Test-Time Training with Self-Supervision for Generalization under Distribution Shifts ICML 2020 Identity Crisis: Memorization and Generalization Under Extreme Overparameterization ICLR 2020 Performative Prediction ICML 2020 The Implicit Fairness Criterion of Unconstrained Learning ICML 2019 Model Similarity Mitigates Test Set Overuse NIPS 2019 A Meta-Analysis of Overfitting in Machine Learning NIPS 2019 Open Problem: How fast can a multiclass test set be overfit? COLT 2019 Stable Recurrent Models ICLR 2019 Delayed Impact of Fair Machine Learning IJCAI 2019 The advantages of multiple classes for reducing overfitting from test set reuse ICML 2019 Natural Analysts in Adaptive Data Analysis ICML 2019 Sanity Checks for Saliency Maps NIPS 2018 Delayed Impact of Fair Machine Learning ICML 2018 Gradient Descent Learns Linear Dynamical Systems JMLR 2018 Avoiding Discrimination through Causal Reasoning NIPS 2017 Train faster, generalize better: Stability of stochastic gradient descent ICML 2016 Equality of Opportunity in Supervised Learning NIPS 2016 The Ladder: A Reliable Leaderboard for Machine Learning Competitions ICML 2015 Generalization in Adaptive Data Analysis and Holdout Reuse NIPS 2015 Differentially Private Learning of Structured Discrete Distributions NIPS 2015 Fast matrix completion without the condition number COLT 2014 Computational Limits for Matrix Completion COLT 2014 The Noisy Power Method: A Meta Algorithm with Applications NIPS 2014 Algorithms and Hardness for Robust Subspace Recovery COLT 2013 A Simple and Practical Algorithm for Differentially Private Data Release NIPS 2012