Discrete Compositional Generation via General Soft Operators and Robust Reinforcement Learning

ICLR 2026 Conference Submission20764 Authors

Keywords: Reinforcement Learning, GFlowNets, Robust RL, Regularized RL, Generative Models, Scientific Discovery

TL;DR: We propose a new general soft operator for discrete compositional generation tasks with experimental promise in various biological sequence design tasks and connect regularized RL/robust RL in these settings.

Abstract: A major bottleneck in scientific discovery consists of narrowing an exponentially large set of objects, such as proteins or molecules, to a small set of promising candidates with desirable properties. While this process can rely on expert knowledge, recent methods leverage reinforcement learning (RL) guided by a proxy reward function to enable this filtering. By employing various forms of entropy regularization, these methods aim to learn samplers that generate diverse candidates that are highly rated by the proxy function. In this work, we make two main contributions. First, we show that these methods are liable to generate overly diverse, suboptimal candidates in large search spaces. To address this issue, we introduce a novel unified operator that combines several regularized RL operators into a general framework that better targets peakier sampling distributions. Secondly, we offer a novel, robust RL perspective of this filtering process. The regularization can be interpreted as robustness to a compositional form of uncertainty in the proxy function (i.e., the true evaluation of a candidate differs from the proxy's evaluation). Our analysis leads us to a novel, easy-to-use algorithm we name trajectory general mellowmax (TGM): we show it identifies higher quality, diverse candidates than baselines in both synthetic and real-world tasks.

Primary Area: generative models

Submission Number: 20764