AVI Architecture: 18 Features Per Variant Fed Into a Linear Hypernetwork Ensemble
anshulkundaje · x · 2026-09-18
Thread segment from Anshul Kundaje on AVI's design: 10 AlphaGenome features are combined with AlphaMissense scores (protein-coding variants), 2 conservation scores, 3 protein loss-of-function annotations, and indel indicators — 18 features per variant. A second model, an ensemble of linear hypernetworks, then maps these diverse variant effect features into a single raw impact score.
More from Research
- Xiaomi MiMo's CodeSpecWeave trains coding agents on interdependent specs for long-horizon coding — XiaomiMiMo · 2026-11-02
- Mistral Large 4 writes a quantum chemistry SCF solver from scratch with one short prompt — MistralAI · 2026-11-02
- Hunting for the most concrete "curing cancer with AI" roadmap of 2025 — iskander · 2026-11-02
- New theory paper proves linear decay is the unique minimax-optimal stepsize schedule for SGD — konstmish · 2026-11-02
- Karpathy's nanochat: the best ChatGPT $100 can buy, near 60k GitHub stars — karpathy · 2026-11-02
- After reading 46 resources on agent evals, one author explains why her 95% score was misleading — blaizedsouza · 2026-11-02