Optimistix: modular optimisation in JAX and Equinox
read the original abstract
We introduce Optimistix: a nonlinear optimisation library built in JAX and Equinox. Optimistix introduces a novel, modular approach for its minimisers and least-squares solvers. This modularity relies on new practical abstractions for optimisation which we call search and descent, and which generalise classical notions of line search, trust-region, and learning-rate algorithms. It provides high-level APIs and solvers for minimisation, nonlinear least-squares, root-finding, and fixed-point iteration. Optimistix is available at https://github.com/patrick-kidger/optimistix.
This paper has not been read by Pith yet.
Forward citations
Cited by 11 Pith papers
-
Differentiable design of the PIAA-ZWFS: a flexible wavefront sensor that approaches the fundamental limit
The PIAA-ZWFS, optimized with differentiable modeling and Bayesian design, closes the gap to the fundamental wavefront sensing limit by a factor of 10 versus conventional ZWFS in photon-limited simulations.
-
RAM: Reachability Across Morphologies
RAM is a morphology-conditioned implicit neural representation trained on 3e10 forward-kinematics samples that serves as a fast, differentiable surrogate for pose reachability and generalizes to unseen morphologies wh...
-
AMIGO: a Data-Driven Calibration of the JWST Interferometer
AMIGO is an end-to-end differentiable forward model of JWST AMI that corrects detector systematics to recover high-precision astrometry and detect close high-contrast companions.
-
Numerical polology: towards next-generation model-building for cosmology
Numerical polology framework samples coupling space to discover ghost-free tensor field theories up to rank three for cosmology, then applies resulting priors to black hole superradiance, dynamical dark energy, and GW data.
-
Efficient High-order Mass-conserving and Energy-balancing Schemes for Schr\"odinger-Poisson Equations
Relaxation techniques applied after each time step to IMEX Runge-Kutta schemes ensure that fully discrete Schrödinger-Poisson systems conserve mass and satisfy energy balance equations up to rounding errors.
-
Accelerated Fourier SAT (AFSAT): Fully Realising a GPU-based Symmetric Pseudo-Boolean SAT Solver
AFSAT realizes FastFourierSAT as a production GPU solver for heterogeneous symmetric pseudo-Boolean SAT via JAX-compiled continuous local search, with tailored DFT for stability and near-linear multi-accelerator scaling.
-
On the boundary cost of source-consistent warp shells
Source-consistent warp shells fail energy conditions at the source-vacuum boundary in all examined constructions and parameter scans.
-
Learning Cross-Coupled and Regime Dependent Dynamics for Aerial Manipulation
A nonlinear latent encoder plus linear latent decoder learns cross-coupled, history-dependent, and nonstationary residual dynamics for aerial manipulators, supporting online Bayesian adaptation and improved real-time ...
-
Complex surface patterning in homo- and heteroepitaxial contexts: (simultaneous) step bunching and step meandering
A linked discrete-continuum model demonstrates that step bunching and meandering can coexist as common growth modes on vicinal surfaces, yielding diverse patterns beyond separate limiting cases.
-
A Study of Parallel Continuous Local Search
Empirical study of parallel continuous local search for SAT finds redundant constraints can slow convergence, CLS works as a hybrid sub-solver, and search stabilizes quickly due to saddle-dense objectives.
-
Curvature-Aware Optimization for High-Accuracy Physics-Informed Neural Networks
Curvature-aware optimizers such as natural gradient and self-scaling BFGS/Broyden accelerate PINN convergence and accuracy on PDEs including Helmholtz, Stokes, Burgers, and Euler equations plus stiff ODEs, with new mo...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.