Research map / Optimization / Optimizers
Differentiable Simulation of Hard Contacts with Soft Gradients for Learning and Control
Anselm Paulus, Andreas René Geist, Pierre Schumacher, Vít Musil, Simon Rappenecker, Georg Martius
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 CFD to MJX (ablation) | MJX · Cube side-length estimation, init 60 mm or 140 mm | convergence | Improved |
| add adaptive integration (DiffMJX) to MJX with smooth collisions (ablation) | MJX · Cube side-length estimation | estimation accuracy | Improved |
| use adaptive-step integration instead of fixed step (ablation) | MJX · cube bounce toy example | loss and gradient error | Improved |
| reduce stepsize in penalty-based simulation (ablation) | penalty-based simulator · point collision toy example | gradient oscillation | Improved |
| reduce stepsize in ideal-elastic simulation (ablation) | ideal-elastic simulator · point collision toy example | gradient sign error | No clear effect |
| add CFD to gradient-based MPC (ablation) | gradient-based MPC (MJX) · bionic tennis | task cost | Improved |
| add CFD to gradient-based MPC (ablation) | gradient-based MPC (MJX) · in-hand ball swap | task cost | No clear effect |
| add CFD to gradient-based MPC (ablation) | gradient-based MPC (MJX) · in-hand cube reorientation | task cost | No clear effect |
| add CFD to gradient-based MPC (ablation) | gradient-based MPC (MJX) · on-the-table cube reorientation | task cost | Improved |
| use gradient-based MPC instead of sampling-based predictive sampling (ablation) | MPC planner · in-hand ball swap and cube reorientation (MyoHand) | task cost | Improved |
| apply CFD naively in the forward pass instead of straight-through (ablation) | CFD · quadruped simulation | simulation realism | Hurt |
| use Tsit5 adaptive solver instead of fixed-step MJX integrators (ablation) | DiffMJX · cube toss | gradient error vs runtime | Improved |
Closest papers by meaning
- DIFFTACTILE: A Physics-based Differentiable Tactile Simulator for Contact-rich Robotic Manipulation (ICLR 2024 · similarity 0.86)
- Geometric Contact Flows: Contactomorphisms for Dynamics and Control (ICML 2025 · similarity 0.86)
- Support-Safe Variational Hybrid Filtering for Contact-Mode and Sparse-Law Recovery (NeurIPS 2026 · similarity 0.84)
- Thin-Shell Object Manipulations With Differentiable Physics Simulations (ICLR 2024 · similarity 0.84)
- Stabilizing Reinforcement Learning in Differentiable Multiphysics Simulation (ICLR 2025 · similarity 0.83)
- Mosaic: A Benchmark Suite for Differentiable Physics Solvers (NeurIPS 2026 · similarity 0.83)
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.
