AReUReDi: Annealed Rectified Updates for Refining Discrete Flows with Multi-Objective Guidance

ICLR 2026 Conference Submission14289 Authors

Keywords: multi-objective optimization, discrete flows, Pareto optimality, biomolecular sequence design, therapeutic peptides, generative modeling

TL;DR: AReUReDi refines discrete flows with annealed updates to guarantee Pareto-optimal multi-objective sequence generation, with strong performance on peptide and SMILES design.

Abstract: Designing sequences that satisfy multiple, often conflicting, objectives is a central challenge in therapeutic and biomolecular engineering. Existing generative frameworks largely operate in continuous spaces with single-objective guidance, while discrete approaches lack guarantees for multi-objective Pareto optimality. We introduce **AReUReDi** (**A**nnealed **Re**ctified **U**pdates for **Re**fining **Di**screte Flows), a discrete optimization algorithm with theoretical guarantees of convergence to the Pareto front. Building on Rectified Discrete Flows (ReDi), AReUReDi combines Tchebycheff scalarization, locally balanced proposals, and annealed Metropolis-Hastings updates to bias sampling toward Pareto-optimal states while preserving distributional invariance. Applied to peptide and SMILES sequence design, AReUReDi simultaneously optimizes up to five therapeutic properties (including affinity, solubility, hemolysis, half-life, and non-fouling) and outperforms both evolutionary and diffusion-based baselines. These results establish AReUReDi as a powerful, sequence-based framework for multi-property biomolecule generation.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 14289