Sparse autoencoders on MolFormer reveal position-tracking latents in early layers and atom-in-substructure plus pharmacologically relevant features in later layers, with non-canonical SMILES causing greater representation disruption than invalid ones.
URL https://www.biorxiv.org/content/ early/2025/06/18/2025.02.06.636901
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
What Does a Chemical Language Model Know About Molecules?
Sparse autoencoders on MolFormer reveal position-tracking latents in early layers and atom-in-substructure plus pharmacologically relevant features in later layers, with non-canonical SMILES causing greater representation disruption than invalid ones.