M-JEPA: Predictive Self-Supervised Learning for Molecular Graphs with Scaffold-Shift Evaluation on Tox21.
M-JEPA (Molecular Joint Embedding Predictive Architectures), a predictive self-supervised method for molecular graphs based on connected-subgraph masking and an exponential-moving-average (EMA) teacher, is evaluated with a compute-matched three-phase protocol that screens objectives on ESOL and tests transfer on Tox21 under Bemis-Murcko scaffold splits.