Hoffman-Yee 2025 continuation grants

The Hoffman-Yee Research Grants are funded by a gift from Reid Hoffman and Michelle Yee to Stanford HAI. The program awards interdisciplinary Stanford teams an initial year of seed funding (approximately $500,000 in year one of the 2024 cohort, with six teams sharing approximately $3 million), followed by competitive continuation grants of up to $2 million per team for two additional years. The April 2025 round announced continuation funding for five of the original six 2024 teams, totaling approximately $10 million for the cohort.

Among the named continuation recipients, Brian Hie, assistant professor of chemical engineering at Stanford, leads the Evo project on open biological foundation models. Evo and its successor Evo 2 are genome-scale foundation models trained to model DNA, RNA, and protein sequences jointly; the Stanford HAI piece describes the line as developing foundation models of genomes that could help decipher human biology and design treatments. The 2024 cohort also included a team of Stanford Law researchers (Daniel Ho, Mark Lemley, Julian Nyarko, plus Megan Ma at the Stanford LIFTLab) working on data creation and attribution for generative AI; whether their grant continued into the 2025 round is not confirmed by the announcement page text.

The program structure (seed-then-continue) is unusual for university AI grant programs. Most internal grant rounds fund a single year at a fixed amount; the Hoffman-Yee model funds long-horizon Stanford research that crosses traditional AI subfield boundaries (CS-plus-law, CS-plus-biology, CS-plus-public-policy), with the year-two-and-three continuation gated on what was produced in the seed year.

In the open-source AI stack the named work spans data, training, weights, and evaluation. The Evo line in particular sits at training and weights as an open biological foundation model release; the law-and-data team is methodologically positioned at the data layer (training-corpus attribution).