Protein language models are trained primarily with masked language modeling (MLM), which predicts masked amino-acid identities.
Joint-embedding predictive architectures (JEPA) instead predict latent representations, but have not been applied to proteins.
ProteinJEPA supplements MLM with a cosine loss for predicting the half-depth hidden states of a teacher given the unmasked sequence.
Performance on 19 Tasks
On 19 tasks, with ESM2 at 35M and 150M parameters and three pretraining seeds, MLM+JEPA outperforms compute-matched and step-matched MLM-only continued training in 78 and 76 of 114 comparisons (14 losses, 22 ties).
The median compute-matched gain is $+0.0106$ on structure- and homology-sensitive tasks versus $+0.0041$ elsewhere, led by SCOPe-40 retrieval and remote homology with improvements of 6.1 percentage points in Recall@1 and 2.7 points in accuracy, respectively.
Gains on these tasks increase with model size from 8M to 150M.
Comparison with Off-the-Shelf Checkpoint
Against the off-the-shelf checkpoint, MLM+JEPA wins 81 of 114 comparisons (median $+0.0068$) without improving MLM loss.
In random initialization the gain is smaller and replicates inconsistently across seeds ($p{=}0.059$).
Improvement of Causal ProGen3 Model
The same recipe improves the causal ProGen3 model, beating a compute-matched next-token-prediction control on 12 of 16 tasks.
Ablations
Ablations show that cosine loss beats mean squared error, while adding shallower targets removes most of the task gain.
JEPA-only training collapses downstream performance: latent prediction complements MLM rather than replacing it.