Differentiate the Solver, Not the Equation: Reverse-Sweep Adjoints for Block Implicit Simulation

• URL: http://arxiv.org/abs/2608.08559v1
• Date: Sun, 09 Aug 2026 08:03:02 GMT
• Title: Differentiate the Solver, Not the Equation: Reverse-Sweep Adjoints for Block Implicit Simulation
• Abstract summary: Differentiable simulation is a key component in learning, control, and inverse problems.<n>We propose solver-level differentiation, which differentiates the executed solver itself.<n>This work highlights solver structure as a practical organizing principle for efficient differentiable simulation.
• Score: 28.75201832533547
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Differentiable simulation is a key component in learning, control, and inverse problems, where gradients through nonlinear implicit solvers are required. Existing approaches either rely on unrolled automatic differentiation, whose memory grows with solver depth, or on equation-level implicit differentiation, which assembles global Jacobians and solves large sparse adjoint systems, discarding the locality of the forward solver -- and differentiating the converged equation rather than the finite computation that actually ran. We propose solver-level differentiation, which differentiates the executed solver itself. When a solver is composed of block implicit updates, its discrete adjoint is obtained by applying the corresponding adjoint updates in reverse order, yielding a reverse-sweep formulation whose backward pass mirrors the forward solver. From an operator perspective, the forward pass realizes an approximate inverse through ordered local solves, and the backward applies its transpose through reverse local adjoint solves, constructing no global system. We instantiate this idea on Vertex Block Descent, yielding a differentiable solver whose reverse colored Gauss-Seidel sweeps are composed entirely of local $3\times 3$ adjoint solves. The backward matches automatic differentiation through the identical executed forward to machine precision at every solver depth, where the equation-level adjoint is off by 37% after one sweep; in a controlled same-codebase, same-GPU comparison it is 33x faster and uses 71x less memory than unrolled automatic differentiation; and the same construction is exact on projective dynamics and extended position-based dynamics. We scale differentiable elastodynamics to $10^6$ contact-coupled soft bodies (8M vertices) on one GPU. Overall, this work highlights solver structure as a practical organizing principle for efficient differentiable simulation.

Related papers

• Time Without Timesteps: Simulating Coupled Dynamical Systems via Self-Consistency [0.0]
  We explore a different formulation for coupled systems.<n>For each subsystem type we train a neural surrogate mapping a full driving trajectory and initial condition directly to a full output trajectory.<n>Following classical waveform relaxation, coupled systems are assembled by enforcing self-consistency among these trajectories.
  arXiv  Detail & Related papers  (2026-09-03T04:32:23Z)
• GRADSOLVE: fast exact gradients for ODE ensembles on GPUs [0.0]
  We present GRADSOLVE, an open-source library for solving and reverse-mode differentiating low-dimensional ODE ensembles on NVIDIA GPUs.<n>It records the steps an adaptive solver accepts and differentiates a fixed-step replay of them; the returned gradient is the exact discrete adjoint of those steps, the same derivative Diffrax returns by default, obtained more cheaply from a fixed-length chain than from an adaptive loop.<n>It targets ensembles differentiated many times against one recorded mesh, keeps Diffrax as a fallback, and supports explicit and Rosenbrock.
  arXiv  Detail & Related papers  (2026-09-02T17:56:10Z)
• Mosaic: A Benchmark Suite for Differentiable Physics Solvers [2.574393697080417]
  Differentiable partial differential equation (PDE) solvers underpin solver-in-the-loop ML training, gradient-based optimal control, and inverse problems.<n>We introduce Mosaic, a benchmarking framework for differentiable PDE solvers that standardizes access to solver gradients.
  arXiv  Detail & Related papers  (2026-06-26T09:38:33Z)
• Closed-Form Spectral Regularization for Multi-Task Model Merging [96.82449201305234]
  Model merging combines several independently fine-tuned experts into a single multi-task model without any training data.<n>State-of-the-art merging methods formulate merging as a layer-wise interference problem.<n>We propose SWUDI, a closed-form method that combines a soft exponential filter, which matches the gradient-flow trajectory of iterative descent.
  arXiv  Detail & Related papers  (2026-06-05T14:00:47Z)
• DInf-Grid: A Neural Differential Equation Solver with Differentiable Feature Grids [73.28614344779076]
  We present a differentiable grid-based representation for efficiently solving differential equations (DEs)<n>Our results demonstrate a 5-20x speed-up over coordinate-based methods, solving differential equations in seconds or minutes while maintaining comparable accuracy and compactness.
  arXiv  Detail & Related papers  (2026-01-15T18:59:57Z)
• A backward differential deep learning-based algorithm for solving high-dimensional nonlinear backward stochastic differential equations [0.6040014326756179]
  We propose a novel backward differential deep learning-based algorithm for solving high-dimensional nonlinear backward differential equations. The deep neural network (DNN) models are trained not only on the inputs and labels but also the differentials of the corresponding labels.
  arXiv  Detail & Related papers  (2024-04-12T13:05:35Z)
• Blending Neural Operators and Relaxation Methods in PDE Numerical Solvers [3.2712166248850685]
  HINTS is a hybrid, iterative, numerical, and transferable solver for partial differential equations. It balances the convergence behavior across the spectrum of eigenmodes by utilizing the spectral bias of DeepONet. It is flexible with regards to discretizations, computational domain, and boundary conditions.
  arXiv  Detail & Related papers  (2022-08-28T19:07:54Z)
• A Unified Framework for Implicit Sinkhorn Differentiation [58.56866763433335]
  We propose an algorithm that obtains analytical gradients of a Sinkhorn layer via implicit differentiation. We show that it is computationally more efficient, particularly when resources like GPU memory are scarce.
  arXiv  Detail & Related papers  (2022-05-13T14:45:31Z)
• Efficient and Modular Implicit Differentiation [68.74748174316989]
  We propose a unified, efficient and modular approach for implicit differentiation of optimization problems. We show that seemingly simple principles allow to recover many recently proposed implicit differentiation methods and create new ones easily.
  arXiv  Detail & Related papers  (2021-05-31T17:45:58Z)
• DiffPD: Differentiable Projective Dynamics with Contact [65.88720481593118]
  We present DiffPD, an efficient differentiable soft-body simulator with implicit time integration. We evaluate the performance of DiffPD and observe a speedup of 4-19 times compared to the standard Newton's method in various applications.
  arXiv  Detail & Related papers  (2021-01-15T00:13:33Z)
• Competitive Mirror Descent [67.31015611281225]
  Constrained competitive optimization involves multiple agents trying to minimize conflicting objectives, subject to constraints. We propose competitive mirror descent (CMD): a general method for solving such problems based on first order information. As a special case we obtain a novel competitive multiplicative weights algorithm for problems on the positive cone.
  arXiv  Detail & Related papers  (2020-06-17T22:11:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.