{"id":"arxiv/2601.00728","url":"https://arxiv.org/abs/2601.00728","tags":["cs.LG","math.NA"],"score":0,"title":"Precision autotuning for linear solvers via contextual bandit-based RL","domain":"cs.LG","source":"arxiv","authors":["Erin Carson","Xinye Chen"],"reviews":0,"subject":"cs.LG","summary":"We propose a reinforcement learning (RL) framework for \\xy{responsive} precision tuning for linear solvers, which can be extended to general algorithms. The framework is formulated as a contextual bandit problem and solved using incremental action-value estimation with a discretized state space to select optimal precision configurations for computational steps, \\xy{retaining} precision and computational efficiency. To verify its effectiveness, we apply the framework to iterative refinement for solving linear systems $Ax = b$. In this application, our approach dynamically chooses precisions…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-01-02","reactions":{}}