Research map / Probabilistic and scientific modeling / Molecules and proteins
Amortized Synthesis of Constrained Configurations Using a Differentiable Surrogate
Xingyuan Sun, Tianju Xue, Szymon Rusinkiewicz, Ryan P Adams
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 amortized synthesis instead of direct supervised learning | Extruder path planning · Extruder path planning test set (simulation) | average Chamfer distance: 0.0314$\pm$0.0008 -> 0.0156$\pm$0.0002 | Improved |
| Use amortized synthesis instead of direct optimization | Extruder path planning · Extruder path planning test set (simulation) | average Chamfer distance: 0.0155$\pm$0.0002 vs. 0.0156$\pm$0.0002 | No clear effect |
| Use amortized synthesis instead of direct supervised learning | Soft robot inverse kinematics · Soft robot inverse kinematics test set | successful cases (over 1,000): 918.3$\pm$3.4 -> 986.3$\pm$0.5 | Improved |
| Use amortized synthesis instead of direct supervised learning | Soft robot inverse kinematics · Soft robot inverse kinematics test set (successful cases) | average distance to target: 0.2171$\pm$0.0016 -> 0.0464$\pm$0.0018 | Improved |
| Use amortized synthesis instead of direct optimization | Soft robot inverse kinematics · Soft robot inverse kinematics test set | successful cases (over 1,000): 986.3$\pm$0.5 vs. 998.3$\pm$0.7 | No clear effect |
| Use amortized synthesis instead of direct optimization | Soft robot inverse kinematics · Soft robot inverse kinematics test set (successful cases) | average distance to target: 0.0233$\pm$0.0003 -> 0.0464$\pm$0.0018 | Hurt |
| Replace the non-linear encoder with a linear encoder (ablation) | Soft robot inverse kinematics · Soft robot inverse kinematics test set | successful cases (over 1,000): 975.0$\pm$3.9 -> 710.7$\pm$24.8 | Hurt |
| Replace the non-linear encoder with a linear encoder (ablation) | Soft robot inverse kinematics · Soft robot inverse kinematics test set (successful cases) | average distance to target: 0.0464$\pm$0.0018 -> 0.1324$\pm$0.0276 | Hurt |
Closest papers by meaning
- Improving Feasibility via Fast Autoencoder-Based Projections (ICLR 2026 · similarity 0.86)
- Co-PiLOT: Constrained Physics-Informed Latent Optimization for Target-Driven Inverse Design (NeurIPS 2026 · similarity 0.84)
- Learning Surrogates for Offline Black-Box Optimization via Gradient Matching (ICML 2024 · similarity 0.83)
- Deep Generative Model for Mechanical System Configuration Design (AAAI 2025 · similarity 0.83)
- StepCAD: Mesh-to-CAD Code Generation via LLM Policy and Geometry-Guided Search (NeurIPS 2026 · similarity 0.82)
- AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction (NeurIPS 2025 · similarity 0.82)
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.
