MIMIC is a split-track encoder-decoder foundation model that unifies sequence reconstruction, prediction, and constrained design across nucleic acids, proteins, and regulatory context using partially observed multimodal inputs.
Maddison, and Bo Wang
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
MIMIC: A Generative Multimodal Foundation Model for Biomolecules
MIMIC is a split-track encoder-decoder foundation model that unifies sequence reconstruction, prediction, and constrained design across nucleic acids, proteins, and regulatory context using partially observed multimodal inputs.