Papers
Papers
* denotes equal contribution
α-β denotes alphabetical order
External Links: Google Scholar. arXiv. Github.
* denotes equal contribution
α-β denotes alphabetical order
External Links: Google Scholar. arXiv. Github.
Evaluating AI Systems: Validity and Reliability
Olawale Salaudeen, Anka Reuel, Ahmed Ahmed, Suhana Bedi, Zachary Robertson, Sudharsan Sundar, Ben Domingue, Angelina Wang, Sanmi Koyejo. Measurement to Meaning: A Validity-Centered Framework for AI Evaluation. Preprint. [arXiv]
Weidinger et al. (incl. Olawale Salaudeen). Toward an Evaluation Science for Generative AI Systems. The Bridge, vol. 55, no. 1, 2025, National Academy of Engineering. [arXiv]
Olawale Salaudeen, Moritz Hardt. ImageNot: A contrast with ImageNet preserves model rankings. In Review. [arXiv] [code]
Robustness-Distribution Shifts: Evaluation, Generalization, and Adaptation
Olawale Salaudeen, Nicole Chiou, Shiny Weng, Sanmi Koyejo. Are Domain Generalization Benchmarks with Accuracy on the Line Misspecified? In Review. [arXiv] [code]
Olawale Salaudeen, Nicole Chiou, Sanmi Koyejo. On Domain Generalization Datasets as Proxy Benchmarks for Causal Representation Learning. Workshop on Causal Representation Learning, Conference on Neural Information Processing Systems (NeurIPS), 2024. (Oral Presentation). [arXiv]
Olawale Salaudeen, Oluwasanmi Koyejo. Causally-Inspired Regularization Enables Domain General Representations. The International Conference on Artificial Intelligence and Statistics (AISTATS), 2024. [arXiv] [code]
Katherine Tsai, Stephen R. Pfohl, Olawale Salaudeen, Nicole Chiou, Matt J. Kusner, Alexander D'Amour, Sanmi Koyejo, Arthur Gretton. Proxy Methods for Domain Adaptation. The International Conference on Artificial Intelligence and Statistics (AISTATS), 2024. [arXiv] [code]
α-β. Alabdulmohsin* et al. (incl. Olawale Salaudeen*). Adapting to Latent Subgroup Shifts via Concepts and Proxies. The International Conference on Artificial Intelligence and Statistics (AISTATS), 2023. [arXiv] [code]
Kusner et al. (incl. Olawale Salaudeen). Adapting to Shifts in Latent Confounders using Observed Concepts and Proxies. Workshop on Principles of Distribution Shift (PODS), International Conference on Machine Learning (ICML), 2022. [paper]
Olawale Salaudeen, Oluwasanmi Koyejo. Exploiting Causal Chains for Domain Generalization. Workshop on Distribution Shifts (DistShift), Conference on Neural Information Processing Systems (NeurIPS), 2021. [paper]
Algorithmic Fairness
Aparna Balagopalan, Kai Wang, Olawale Salaudeen, Asia Biega, Marzyeh Ghassemi. What’s in a Query: Polarity-aware Distribution-based Fair Ranking. The International World Wide Web Conference (WWW), 2025. [arXiv]
Pfohl et al. (incl. Olawale Salaudeen). Understanding subgroup performance differences of fair predictors using causal models. Workshop on Distribution Shifts (DistShift), Conference on Neural Information Processing Systems (NeurIPS), 2023. [paper]
Chirag Nagpal, Olawale Salaudeen, Sanmi Koyejo, Stephen Pfohl. Addressing Observational Biases in Algorithmic Fairness Assessments. Workshop on Algorithmic Fairness through the Lens of Causality and Privacy, Conference on Neural Information Processing Systems (NeurIPS), 2022 (extended abstract). [poster]
Neuroscience and Neuroimaging
Olawale Salaudeen, Paul Camacho, Aron Barbey, Brad Sutton, Sanmi Koyejo. Enhancing fMRI Motion Denoising with ICA-AROMA and Causal Discovery. In Review. [code]
Sutton et al. (incl. Olawale Salaudeen). Ultra-fast 3D fMRI to explore cardiac-induced fluctuations in BOLD-based functional imaging. International Society for Magnetic Resonance in Medicine (ISMRM), 2022 (abstract). [link]