Research map / 3D vision / 3D generation
LagNet: Deep Lagrangian Mechanics for Plug-and-Play Molecular Representation Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| add the Lagrangian module (ablation) | LagNet · Tox21 | Multi-ROC: 0.856 -> 0.867 | Helped |
| add the Lagrangian module (ablation) | LagNet · BBBP | ROC: 0.952 -> 0.961 | Helped |
| add the Lagrangian module (ablation) | LagNet · Lipop | RMSE: 0.551 -> 0.538 | Helped |
| add the Lagrangian module (ablation) | LagNet · FreeSolv | RMSE: 0.789 -> 0.761 | Helped |
| add the Lagrangian module (ablation) | LagNet · ESOL | RMSE: 0.558 -> 0.541 | Helped |
| replace the 3-hop loss with 1-hop loss (ablation) | LagNet · QM9 | Kabsch-RMSD (Å): 1.180 -> 0.996 | Helped |
| replace the 3-hop loss with 1-hop loss (ablation) | LagNet · QM9 | Distance Loss (10−2 Å): 0.201 -> 0.181 | Helped |
| replace the 3-hop loss with 1-hop loss (ablation) | LagNet · QM9 | 3-hop Loss (103 Å): 0.402 -> 1.742 | Hurt |
| use LagNet rather than the strongest reported baseline | molecular property prediction · Tox21 | Multi-ROC: 0.903 -> 0.867 | Hurt |
| use LagNet rather than the strongest reported baseline | molecular property prediction · BBBP | ROC: 0.960 -> 0.961 | Helped |
| use LagNet rather than the strongest reported baseline | molecular property prediction · Lipop | RMSE: 0.541 -> 0.538 | Helped |
| use LagNet rather than the strongest reported baseline | molecular property prediction · FreeSolv | RMSE: 0.767 -> 0.761 | Helped |
Closest papers by meaning
- Conformation-Guided Molecular Representation with Hamiltonian Neural Networks (ICLR 2021 · similarity 0.89)
- Learning Over Molecular Conformer Ensembles: Datasets and Benchmarks (ICLR 2024 · similarity 0.88)
- A Plug-and-Play Quaternion Message-Passing Module for Molecular Conformation Representation (AAAI 2024 · similarity 0.87)
- SubgDiff: A Subgraph Diffusion Model to Improve Molecular Representation Learning (NeurIPS 2024 · similarity 0.87)
- Swallowing the Bitter Pill: Simplified Scalable Conformer Generation (ICML 2024 · similarity 0.87)
- BindGPT: A Scalable Framework for 3D Molecular Design via Language Modeling and Reinforcement Learning (AAAI 2025 · similarity 0.86)
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.