GenomAI v2: Multimodal Foundation Model for Genomic Sequences

📄 Research Paper

GenomAI_v2_Research_Paper.md

🔬 Key Innovations

• Contrastive Self-Supervised Learning
• Discrete Diffusion Models
• Cross-Modal Attention

📊 Results

• 0.5655 accuracy on population classification
• 734K parameter transformer
• T4 GPU training in <5 minutes