Same Physics, Different Gradients: A State-of-the-Art Report on Differentiable Physics Simulation
Same Physics, Different Gradients: A State-of-the-Art Report on Differentiable Physics Simulation
This paper presents a comprehensive report on differentiable physics simulation, separating the literature's conflated questions about which mathematical object is differentiated and how the backward pass is organized.
This publication provides a taxonomy of differentiable physics simulation methods, distinguishing between the mathematical object differentiated (referent) and the backward pass implementation (realization), while analyzing gradient loss mechanisms in non-smooth contact scenarios.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
arXiv:2611.12162v1 Announce Type: cross Abstract: Differentiable simulation propagates derivatives through a physics simulator, turning animation, system identification, control, and computational design into gradient-based optimization. Engines differ in the physics they model and, less visibly, in which mathematical object they differentiate. We separate two questions the literature conflates. The referent is which
Open source S2
Source：arXiv Robotics — research abstracts · arxiv.org