Community ports Illumina's PromoterAI genomics model to PyTorch with new optimizations
jmschreiber91 · x · 2026-11-12
A developer has ported Illumina's PromoterAI v1 — a deep learning model predicting the regulatory impact of promoter DNA variants on gene expression — from TensorFlow/Keras to PyTorch, releasing it as promoterai-torch on GitHub under MIT license with two major new optimizations. The project includes examples, tests and a companion blog post, and is an unofficial community effort; the original codebase uses a restrictive PolyForm Strict license.
More from Research
- LeapQuant: training-free quantization makes linear attention up to 4.25x faster — tianyin_xu · 2026-11-12
- K-means for supervised learning: self-supervised learning as a special case — caglar_ee · 2026-11-12
- SPoT boosts calibration: AUROC 0.61 to 0.84 while learning to abstain — polkirichenko · 2026-11-12
- Meta researchers unveil SPoT post-thinking: +17.7pp safety, +4.7pp math accuracy over RL baseline — polkirichenko · 2026-11-12
- Sokoban Eval lets the crowd race LLMs for shortest solutions across 66 rooms — banteg · 2026-11-12
- 1920 Popular Mechanics clipping shows eclipse that confirmed light bends under gravity — generativist · 2026-11-12