TMol puts Rosetta-style all-atom energies in PyTorch, 350x throughput on batched H200s
ArashVahdat · x · 2026-11-20
- TMol (paper on arXiv) brings Rosetta-style all-atom molecular energetics into PyTorch with GPU batching and gradients, letting you embed molecular energetics directly into deep learning models.
- Benchmark: >350x higher scoring + gradient throughput with batched H200 GPUs vs single-core PyRosetta.
- The workflow closes the loop: AI generates biomolecular designs, energy scores assess predicted structures, and search refines candidates — physics-guided AI for digital biology.
- Related to the RFD4-Proteina work; detailed in Kieran Didi's thread.
More from Research
- Roboticist: needing ~10 future command frames for joint tracking signals a flawed formalism — KyleMorgenstein · 2026-11-20
- Road MOPD: resolving multi-teacher distillation conflicts, +5.6–8.8 pts on medical tasks — jiqizhixin · 2026-11-20
- Roboticist argues full joint space is the wrong abstraction for high-level agents — KyleMorgenstein · 2026-11-20
- Do language models have explicit goal encodings? New interpretability paper takes first step — benno_krojer · 2026-11-20
- Regularized Score Centering fixes advantage estimation in agentic RL, 9.9× on BabyAI — aviral_kumar2 · 2026-11-20
- DeepMind Adds Memory to Recursive Self-Improvement, Boosting Scores 68→85 — Dr_Singularity · 2026-11-20