publications
2025
- NeurIPSPin the Tail on the Model: Blindfolded Repair of User-Flagged Failures in Text-to-Image ServicesIn Advances in Neural Information Processing Systems, 2025The first two authors contributed equally to this work.
2024
- ESORICSPrivacy-Preserving Regular Expression Matching using TNFAIn ESORICS 2024: 29th European Symposium on Research in Computer Security, 2024
- ICLRConfidential-DPproof: Confidential Proof of Differentially Private TrainingIn ICLR 2024: The Twelfth International Conference on Learning Representations, 2024Spotlight Paper (top 5%)
2023
- VLDBRESCU-SQL: Oblivious Querying for the Zero Trust CloudProceedings of the VLDB Endowment, 2023The first two authors contributed equally to this work.
2022
- ICLRReproducible Subjective EvaluationIn ICLR Workshop on Machine Learning Evaluation Standards, 2022