Research map / Image generation / Autoregressive and latent generation
Beyond Conservation: Flexible Molecular Assembly with Unbalanced Diffusion Bridge
Rongchao Zhang, Yiwei Lou, Yu Huang, Yi Xin, Yongzhi Cao, Hanpin Wang
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 |
|---|---|---|---|
| add unbalanced strategy (ablation) | AssemUDB · Cluster-5k | PMS (center): 6.12 -> 5.93 | Improved |
| add unbalanced strategy (ablation) | AssemUDB · Cluster-5k | PMS (atom): 7.18 -> 6.97 | Improved |
| add unbalanced strategy (ablation) | AssemUDB · Cluster-5k | Separation: 96.29 -> 98.08 | Improved |
| add unbalanced strategy (ablation) | AssemUDB · Cluster-5k | Collision: 0.35 -> 0.34 | No clear effect |
| use full attention/head layers instead of 6 (ablation) | AssemUDB · Cluster-5k | PMS (center): 6.08 -> 5.93 | Improved |
| use full attention/head layers instead of 6 (ablation) | AssemUDB · Cluster-5k | PMS (atom): 7.06 -> 6.97 | Improved |
| use full attention/head layers instead of 6 (ablation) | AssemUDB · Cluster-5k | Separation: 97.08 -> 98.08 | Improved |
| use AssemUDB instead of AssembleFlow | AssembleFlow · Cluster-5k | PMS (center): 6.13 -> 5.93 | Improved |
| use AssemUDB instead of AssembleFlow | AssembleFlow · Cluster-113k | Separation: 98.15 -> 95.97 | Hurt |
| use AssemUDB instead of PackMol | PackMol · Cluster-5k | PMS (center): 6.05 -> 5.93 | Improved |
| use AssemUDB instead of PackMol | PackMol · Cluster-5k | Separation: 99.56 -> 98.08 | Hurt |
| use AssemUDB instead of PackMol | PackMol · Cluster-113k | Separation: 99.42 -> 95.97 | Hurt |
Closest papers by meaning
- Energy-Based Flow Matching for Generating 3D Molecular Structure (ICML 2025 · similarity 0.88)
- Generative Molecular Morphing for Flexible-Size Design via Unbalanced Optimal Transport (NeurIPS 2026 · similarity 0.88)
- Swallowing the Bitter Pill: Simplified Scalable Conformer Generation (ICML 2024 · similarity 0.87)
- Unified Guidance for Geometry-Conditioned Molecular Generation (NeurIPS 2024 · similarity 0.87)
- DiSCO: Diffusion Schrödinger Bridge for Molecular Conformer Optimization (AAAI 2024 · similarity 0.87)
- SE(3)-Stochastic Flow Matching for Protein Backbone Generation (ICLR 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.
