AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

May the Force be with You: Unified Force-Centric Pre-Training for 3D Molecular Conformations

Rui Feng, Qi Zhu, Huan Tran, Binghong Chen, Aubrey Toland, Rampi Ramprasad, Chao Zhang

NeurIPS 2023 · Architectures and efficiency › RNNs and normalization · approach: Scalable approximation and equivariant compute · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add zero-force and de-noising regularizations (ablation)ET-ORE · MD17 Aspirinforce MAE: 4.2 -> 1.0Improved
add zero-force and de-noising regularizations (ablation)ET-ORE · MD17 Aspirinstability (ps): 94 -> 300Improved
add zero-force and de-noising regularizations (ablation)ET-ORE · MD17 Naphthalenestability (ps): 25 -> 300Improved
add zero-force and de-noising regularizations (ablation)ET-ORE · MD17 Ethanolh(r) MAE: 0.10 -> 0.03Improved
add zero-force and de-noising regularizations (ablation)ET-ORE · MD17 Ethanolstability (ps): 300 -> 300No clear effect
pre-train with forces only (ET-ORE) instead of training from scratch (ablation)TorchMDNet · MD17 Aspirinforce MAE: 7.4 -> 4.2Improved
pre-train with forces only (ET-ORE) instead of training from scratch (ablation)TorchMDNet · MD17 Aspirinstability (ps): 102 -> 94No clear effect
full force-centric pre-training before fine-tuning (ablation)TorchMDNet · MD17 Aspirinforce MAE: 7.4 -> 1.0Improved
full force-centric pre-training before fine-tuning (ablation)TorchMDNet · MD17 Naphthalenestability (ps): 50 -> 300Improved
pre-train with force-centric objective on diverse conformations (ablation)TorchMDNet · QM9 HOMOMAE (meV): 20.3 -> 16.8Improved
pre-train with force-centric objective on diverse conformations (ablation)TorchMDNet · QM9 HOMO-LUMO gapMAE (meV): 36.1 -> 26.4Improved
force-supervised off-equilibrium pre-training instead of de-noising pre-trainingNoisyNode · QM9 HOMO-LUMO gapMAE (meV): 24.5 -> 26.4Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy