Lindorff-Larsen argues current MLIPs lag classical force fields on biomolecular accuracy
LindorffLarsen · x · 2026-10-28
KTH professor Erik Lindorff-Larsen and MLIP researcher Timothy Duignan debated on X over machine-learning interatomic potentials vs classical force fields. Lindorff-Larsen says he hasn't seen evidence that current MLIPs approach classical FF accuracy on biomolecular problems; Duignan notes ΩMol-trained models only just started working well (a vv10 issue was recently fixed) and asks what nanosecond/10k-atom benchmarks would be convincing. Lindorff-Larsen replies that a bottom-up strategy may work, but current QM data types are insufficient and MLIPs lack benchmarks comparable to a decade of classical FF validation.
Related event: Researchers Debate Whether MLIPs Can Outperform Classical Force Fields(5 posts)→
More from Research
- DeepMind researcher: all of mathematics will soon be formalizable, with agents hunting proofs — ChrSzegedy · 2026-10-28
- A fine-tuned 'cat obsession' control fools every AI welfare criterion — MirrorEthic_Anchor · 2026-10-28
- Sweeping every Qwen3-8B neuron, rated for weirdness by Haiku — Sauers_ · 2026-10-28
- Kevin Buzzard: Lean formalization revealed errors in theorems by 4 Fields medalists — Machine Learning Street Talk · 2026-10-28
- Follow-up repo machine-verifies Hamiltonian cycle results on top of OpenAI/Anthropic ETH refutation claims — xennygrimmato_ · 2026-10-28
- DiscoTrace wins EMNLP 2026 Best Paper: LLM answers lack humans' community-specific rhetorical diversity — TuhinChakr · 2026-10-28