Technical University of Munich

Protein-family language model trained on unaligned homolog sets for zero-shot variant fitness prediction and design. ProFam-1 holds 251M parameters.

2B-parameter medical vision-language model that uses reinforcement learning to show interpretable reasoning for radiology visual question answering.

Masked DNA language model trained on over 800 vertebrate genomes and conditioned on species identity to learn conserved regulatory sequence features.