CSAIL MIT Postdoc.
Working on learning generative models from corrupted data.
CSAIL MIT Postdoc.
Working on learning generative models from corrupted data.
[NeurIPS 2023] Official Implementation: "Ambient Diffusion: Learning Clean Distributions from Corrupted Data"
[ICML 2024]: Official implementation for the paper: "Consistent Diffusion Meets Tweedie"
💫 Industrial-strength Natural Language Processing (NLP) in Python