Research map / Probabilistic and scientific modeling / Molecules and proteins
Coupled Cluster con MoLe: Molecular Orbital Learning for Neural Wavefunctions
Luca Thiede, Abdulrahman Aldossary, Andreas Burger, Jorge A. Campos-Gonzalez-Angulo, Alex Zook, Melisa Alkan, KOHEI NAKAJI, Taylor Lee Patti, Jérôme Florian Gonthier, Mohammad Ghazi Vakili, Alan Aspuru-Guzik
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use localized orbitals instead of canonical orbitals (ablation) | MoLe (64 irreps) · QM7 | energy MAE (mHa): 0.27 -> 0.16 | Improved |
| predict CCSD amplitudes with MoLe instead of MACE trained directly on energy | energy models · QM7 (full train) | energy MAE (mHa): 1.83 -> 0.12 | Improved |
| use MoLe instead of Delta-MP2 MACE with 100 training molecules | MACE (Delta-MP2) · QM7 (100 molecules) | energy MAE (mHa): 1.64 -> 0.66 | Improved |
| use MoLe instead of Delta-MP2 MACE with 100 training molecules | MACE (Delta-MP2) · PubChem (100 molecules) | energy MAE (mHa): 5.29 -> 2.80 | Improved |
| use MoLe instead of Delta-MP2 MACE with full training set | MACE (Delta-MP2) · Amino acids | energy MAE (mHa): 0.49 -> 0.78 | Hurt |
| use MoLe instead of eSEN Delta-MP2 with full training set | eSEN (Delta-MP2) · Diels-Alder | energy MAE (mHa): 1.15 -> 1.16 | No clear effect |
| use MoLe amplitudes as CCSD initial guess instead of MP2 | CCSD solver · PubChem | average cycles: 10.11 -> 6.3 | Improved |
| use MoLe density matrix instead of MP2 | 1-RDM · Amino acids | Frobenius error: 0.59 -> 0.13 | Improved |
| increase transformer depth 1 to 4 layers and width 32 to 128 (ablation) | MoLe · QM7 | T2 error | Improved |
Closest papers by meaning
- Enhancing the Scalability and Applicability of Kohn-Sham Hamiltonians for Molecular Systems (ICLR 2025 · similarity 0.89)
- Gaussian Plane-Wave Neural Operator for Electron Density Estimation (ICML 2024 · similarity 0.87)
- EDBench: Large-Scale Electron Density Data for Molecular Modeling (NeurIPS 2025 · similarity 0.87)
- Learning the Electronic Hamiltonian of Large Atomic Structures (ICML 2025 · similarity 0.87)
- A Recipe for Charge Density Prediction (NeurIPS 2024 · similarity 0.87)
- Higher-Rank Irreducible Cartesian Tensors for Equivariant Message Passing (NeurIPS 2024 · similarity 0.87)
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.
