Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.26743

FOMTools: Development and AI-Assisted Code Generation for First-Order Methods

Vinit Ranjan·Daniel Deza·Anushka Acharya·Jisun Park·Bartolomeo Stellato
View original source ↗Read source document ↗Export metadata ↓

Abstract

We present FOMTools, an open-source Python package for developing, analyzing, and deploying first-order algorithms for parametric linear and convex quadratic optimization. The user specifies an iteration in Python, with syntax close to its mathematical description. The package represents the iteration as a computation graph of affine and piecewise affine operations. To analyze worst-case performance, FOMTools formulates a mixed-integer verification problem that maximizes a convergence residual over a specified set of problem instances and initial iterates. For probabilistic performance analysis, a batched GPU simulator collects algorithm trajectories and computes statistical bounds on the probability of exceeding a specified convergence tolerance. From the same algorithm description, FOMTools generates specialized solver code for embedded hardware, multicore CPUs, and GPUs. The package first generates the numerical data, factorizations, and matrix-vector routines without any agent. A coding agent then translates the iteration into C or CUDA code and iteratively optimizes the implementation. After each phase, the package runs the generated code on sampled inputs and compares it with a reference execution. FOMTools also serves as a CVXPYgen backend, generating and optimizing solvers for parametric CVXPY problems. We demonstrate FOMTools on sparse coding, network optimization, optimal control, and portfolio rebalancing. The verification bounds reveal worst-case behavior that sampling misses. The generated solvers run up to two orders of magnitude faster than NumPy implementations, and they outperform hand-written GPU code and the OSQP solvers generated by CVXPYgen.

Explore related subjects

math.OC

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Vinit Ranjan, Daniel Deza, Anushka Acharya, Jisun Park, Bartolomeo Stellato. 2026-10-19. FOMTools: Development and AI-Assisted Code Generation for First-Order Methods. https://arxiv.org/abs/2610.26743

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

The Principle of Optimality in Dynamic Programming: A Pedagogical Note

The principle of optimality is a fundamental aspect of dynamic programming, which states that the optimal solution to a dynamic optimization problem can be found by combining the optimal solutions to its sub-problems. While this principle is generally applicable, it is often only taught for problems with finite or countable state spaces in order to sidestep measure-theoretic complexities. Therefore, it cannot be applied to classic models such as inventory management and dynamic pricing models that have continuous state spaces, and students may not be aware of the possible challenges involved in studying dynamic programming models with general state spaces. To address this, we provide conditions and a self-contained simple proof that establish when the principle of optimality for discounted dynamic programming is valid. These conditions shed light on the difficulties that may arise in the general state space case. We provide examples from the literature that include the relatively involved case of universally measurable dynamic programming and the simple case of finite dynamic programming where our main result can be applied to show that the principle of optimality holds.

math.OC↗
arXiv2026

A Geometric Framework for Stochastic Iterations

This paper concerns models and convergence principles for dealing with stochasticity in a wide range of algorithms arising in nonlinear analysis and optimization in Hilbert spaces. It proposes a flexible geometric framework within which existing solution methods can be recast and improved, and new ones can be designed. Almost sure weak, strong, and linear convergence results are established in particular for stochastic fixed point iterations, the stochastic gradient descent method, and stochastic extrapolated parallel algorithms for feasibility problems. In these areas, the proposed algorithms exceed the features and convergence guarantees of the state of the art. Numerical applications to signal and image recovery are provided.

math.OC↗
arXiv2026

Attention's forward pass and Frank-Wolfe

We study the hardmax limit of self-attention dynamics for token embeddings obtained in the zero-temperature ($β\to+\infty$) regime, and relate it to the finite-$β$ setting. In this limit, the update rule can be viewed as a Frank-Wolfe step for a quadratic objective over the convex hull of the current token embeddings. When the key-query matrix is negative semidefinite, the method linearly contracts all tokens to a single cluster at the origin. When it is positive semidefinite, extending the hardmax rule to the entire convex hull induces a Voronoi diagram: vertices are stationary, interior points remain in their initial cells, and each token moves along a straight line toward its cell's vertex, yielding (super-)exponential convergence. As a byproduct, we also establish well-posedness of the associated ODE limit in this regime. Returning to the finite-$β$ regime, we model self-attention dynamics as a Markov chain and prove dynamic metastability: with high probability, interior tokens reach near-vertex configurations in a constant number of steps and remain within a small neighborhood for times that grow exponentially in the inverse temperature $β$, before ultimately collapsing to the origin. Thus, the hardmax dynamics accurately approximate the finite-$β$ process over exponentially long time horizons.

math.OC↗