Flow-based Extremal Mathematical Structure Discovery

• URL: http://arxiv.org/abs/2601.18005v1
• Date: Sun, 25 Jan 2026 21:41:47 GMT
• Title: Flow-based Extremal Mathematical Structure Discovery
• Authors: Gergely Bérczi, Baran Hashemi, Jonas Klüver,
• Abstract summary: FlowBoost is a closed-loop generative framework that learns to discover rare and extremal geometric structures.<n>We demonstrate the framework on four geometric optimization problems.
• Score: 1.1087735229999816
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The discovery of extremal structures in mathematics requires navigating vast and nonconvex landscapes where analytical methods offer little guidance and brute-force search becomes intractable. We introduce FlowBoost, a closed-loop generative framework that learns to discover rare and extremal geometric structures by combining three components: (i) a geometry-aware conditional flow-matching model that learns to sample high-quality configurations, (ii) reward-guided policy optimization with action exploration that directly optimizes the generation process toward the objective while maintaining diversity, and (iii) stochastic local search for both training-data generation and final refinement. Unlike prior open-loop approaches, such as PatternBoost that retrains on filtered discrete samples, or AlphaEvolve which relies on frozen Large Language Models (LLMs) as evolutionary mutation operators, FlowBoost enforces geometric feasibility during sampling, and propagates reward signal directly into the generative model, closing the optimization loop and requiring much smaller training sets and shorter training times, and reducing the required outer-loop iterations by orders of magnitude, while eliminating dependence on LLMs. We demonstrate the framework on four geometric optimization problems: sphere packing in hypercubes, circle packing maximizing sum of radii, the Heilbronn triangle problem, and star discrepancy minimization. In several cases, FlowBoost discovers configurations that match or exceed the best known results. For circle packings, we improve the best known lower bounds, surpassing the LLM-based system AlphaEvolve while using substantially fewer computational resources.

Related papers

• Riemannian MeanFlow [28.880123107817525]
  We introduce a framework for learning flow maps directly on inference, enabling high-quality generations with as few as one forward pass.<n>We show that few-step flow maps enable efficient reward-guided design through reward look-ahead, where terminal states can be predicted from intermediate steps.
  arXiv  Detail & Related papers  (2026-02-08T00:41:17Z)
• Predictive Feature Caching for Training-free Acceleration of Molecular Geometry Generation [67.20779609022108]
  Flow matching models generate high-fidelity molecular geometries but incur significant computational costs during inference.<n>This work discusses a training-free caching strategy that accelerates molecular geometry generation.<n> Experiments on the GEOM-Drugs dataset demonstrate that caching achieves a twofold reduction in wall-clock inference time.
  arXiv  Detail & Related papers  (2025-10-06T09:49:14Z)
• Refine Drugs, Don't Complete Them: Uniform-Source Discrete Flows for Fragment-Based Drug Discovery [0.0]
  We introduce InVirtuoGen, a discrete flow generative model for fragmented SMILES for de novo and fragment-constrained generation.<n>For property and lead optimization, we propose a hybrid scheme that combines a genetic algorithm with a Proximal Property Optimization fine-tuning strategy.<n>Our approach sets a new state-of-the-art on the Practical Molecular Optimization benchmark, measured by top-10 AUC across tasks.
  arXiv  Detail & Related papers  (2025-09-30T15:34:53Z)
• Learnable SMPLify: A Neural Solution for Optimization-Free Human Pose Inverse Kinematics [13.621560002904873]
  Learnable SMPLify is a neural framework that replaces the iterative fitting process in SMPLify with a single-pass regression model.<n>It achieves nearly 200x faster runtime compared to SMPLify, generalizes well to unseen 3DPW and RICH, and operates as a model-agnostic manner when used as a plug-in tool on LucidAction.
  arXiv  Detail & Related papers  (2025-08-19T06:53:57Z)
• Training Free Guided Flow Matching with Optimal Control [6.729886762762167]
  We present OC-Flow, a training-free framework for guided flow matching using optimal control.<n>We show that OC-Flow achieved superior performance in experiments on text-guided image manipulation, conditional molecule generation, and all-atom peptide design.
  arXiv  Detail & Related papers  (2024-10-23T17:53:11Z)
• A Sample Efficient Alternating Minimization-based Algorithm For Robust Phase Retrieval [56.67706781191521]
  In this work, we present a robust phase retrieval problem where the task is to recover an unknown signal. Our proposed oracle avoids the need for computationally spectral descent, using a simple gradient step and outliers.
  arXiv  Detail & Related papers  (2024-09-07T06:37:23Z)
• Flow Matching on General Geometries [43.252817099263744]
  We propose a simple yet powerful framework for training continuous normalizing flows on manifold geometries. We show that it is simulation-free on simple geometries, does not require divergence, and computes its target vector field in closed-form. Our method achieves state-of-the-art performance on many real-world non-Euclidean datasets.
  arXiv  Detail & Related papers  (2023-02-07T18:21:24Z)
• Faster Adaptive Federated Learning [84.38913517122619]
  Federated learning has attracted increasing attention with the emergence of distributed data. In this paper, we propose an efficient adaptive algorithm (i.e., FAFED) based on momentum-based variance reduced technique in cross-silo FL.
  arXiv  Detail & Related papers  (2022-12-02T05:07:50Z)
• Deep Shells: Unsupervised Shape Correspondence with Optimal Transport [52.646396621449]
  We propose a novel unsupervised learning approach to 3D shape correspondence. We show that the proposed method significantly improves over the state-of-the-art on multiple datasets.
  arXiv  Detail & Related papers  (2020-10-28T22:24:07Z)
• AdaLead: A simple and robust adaptive greedy search algorithm for sequence design [55.41644538483948]
  We develop an easy-to-directed, scalable, and robust evolutionary greedy algorithm (AdaLead) AdaLead is a remarkably strong benchmark that out-competes more complex state of the art approaches in a variety of biologically motivated sequence design challenges.
  arXiv  Detail & Related papers  (2020-10-05T16:40:38Z)
• Learning to Optimize Non-Rigid Tracking [54.94145312763044]
  We employ learnable optimizations to improve robustness and speed up solver convergence. First, we upgrade the tracking objective by integrating an alignment data term on deep features which are learned end-to-end through CNN. Second, we bridge the gap between the preconditioning technique and learning method by introducing a ConditionNet which is trained to generate a preconditioner.
  arXiv  Detail & Related papers  (2020-03-27T04:40:57Z)

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.