Across 13 protein language models and 15 downstream tasks, the last layer is rarely the best source of embeddings; intermediate layers work better, and the best layer is set by dataset structure rather than by the task alone.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Task- and dataset-specific information in protein language models
Across 13 protein language models and 15 downstream tasks, the last layer is rarely the best source of embeddings; intermediate layers work better, and the best layer is set by dataset structure rather than by the task alone.