Hey, I’m a PhD Candidate at MIT. My research focuses on the intersection of AI and policy: responsibly training, evaluating, and governing general-purpose AI systems. I lead the Data Provenance Initiative, led the Open Letter on A Safe Harbor for Independent AI Evaluation & Red Teaming, and have contributed to training models like Bloom, Aya, and Flan-T5/PaLM. I’m thankful for the recognition my research has received: Best Paper Awards from ACL 2024, NAACL 2024, as well as coverage by the NYT, Washington Post, Atlantic, 404 Media, Vox, and MIT Tech Review.
Prior:
2024.08: Aya Model wins Best Paper Award at ACL 2024.
2024.08: Consent in Crisis covered by the NYT, 404 Media, Vox, and Yahoo! Finance.
2024.07: A Pretrainer’s Guide to Training Data wins Outstanding Paper Award at NAACL 2024.
2024.07: 3 Oral and 1 Spotlight paper accepted to ICML 2024: (1) Safe Harbor, (2) Societal Impact of Open Foundation Models, (3) AI Autonomous Weapons Risk Geopolitical Instability, and (4) Data Authenticity, Consent, and Provenance for AI Are All Broken: What Will It Take to Fix Them?.
2024.06: The Data Provenance Initiative was awarded the Mozilla Data Futures Lab grant and wins the MIT Generative AI Impact Award, funded for $70,000. Presented at MozFest 2024.
2024.05: Co-wrote the International Scientific Report on the Safety of Advanced AI.
2024.03: Our Open Letter on A Safe Harbor for Independent AI Evaluation & Red Teaming garnered 350+ signatures from leading researchers. Covered by the Washington Post, VentureBeat, and the Knight First Amendment Institute at Columbia University. Cited in the US Department of Justice’s letter to the US Copyright Office.
2023.10: Launched the The Data Provenance Initiative, covered by the Washington Post, VentureBeat, and IEEE Spectrum.
2023.09: New paper on Foundation Model Transparency Index, covered by NYT, The Atlantic, and VentureBeat.
2023.02-06: Co-instructor for MIT’s Generative AI course MAS.S68.