Researchers debate whether MLIPs beat classical force fields for protein simulation
TimothyDuignan · x · 2026-10-28
A substantive debate on machine-learned interatomic potentials (MLIPs) vs classical force fields:
- Timothy Duignan argues classical force fields can't even reproduce sodium chloride activity coefficients without extensive experimental fitting — empirical parameterization that doesn't scale.
- Lindorff Larsen counters that while the latest MLIPs make nanosecond, quantum-accuracy protein simulations trivial, there's no evidence yet they beat classical force fields on standard biomolecular problems, and ns-scale simulations were possible 30 years ago.
- Duignan adds that hand-fitting zinc coordination (tetrahedral vs octahedral) remains unsolved, and classical force fields could still improve rapidly and win in many applications.
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