Logan Engstrom
Source-listed: Graduate Student · Massachusetts Institute of Technology (MIT), Computer Science and Artificial Intelligence Laboratory (CSAIL)
Record snapshot: Sep 3, 2026. Coverage and affiliations may be incomplete or historical.
Selected work
5- Optimizing ML Training with Metagradient DescentSaved credit: Contributor · 2025Explore people connected to this work →
- Datamodels: Predicting Predictions from Training DataSaved credit: Contributor · 2022Explore people connected to this work →
- FFCV: Accelerating Training by Removing Data BottlenecksSaved credit: Contributor · 2023Explore people connected to this work →
- Implementation Matters in Deep Policy Gradients: A Case Study on PPO and TRPOSaved credit: Contributor · 2020Explore people connected to this work →
- Adversarial Examples Are Not Bugs, They Are FeaturesSaved credit: Contributor · 2019Explore people connected to this work →