Scaling unlocks broader generation and deeper functional understanding of proteins
Aadyot Bhatnagar, Sarthak Jain, Joel Beazer, Samuel C. Curran, Alexander M. Hoffnagle, Kyle Ching, Michael Martyn, Stephen Nayfach, Jeffrey A. Ruffolo, Ali Madani
Scaling sparse Mixture-of-Experts protein language models up to 46B parameters on 1.5T curated tokens broadens viable protein generation across families and amplifies gains from laboratory preference alignment.
neural information processing systems · Jan 1, 2025Peer-reviewedSynthetic Biology✦ In-depth digest