Mol-JEPA: Molecular Foundation Model Learning Across 14 Modalities

Boehringer Ingelheim and partners introduced Mol-JEPA, a molecular foundation model based on a multimodal joint-embedding predictive architecture that learns from 14 modalities such as structure, cell imaging and ADMET. The paper on arXiv reports state-of-the-art results on multiple benchmarks.

2026-08-26 ~ 2026-08-26 · 2 related posts

• Mol-JEPA: Molecular Foundation Model Learns from 14 Modalities Simultaneously — rbhar90 · 2026-08-26
• Mol-JEPA: multimodal joint-embedding predictive architecture hits SOTA on molecule benchmarks — randall_balestr · 2026-08-26