Helix: Evolutionary Reinforcement Learning for Open-Ended Scientific Problem Solving

ICLR 2026 Conference Submission20166 Authors

Keywords: Large Language Models, Reinforcement Learning, Evolution Strategies, Scientific Discovery

TL;DR: We introduce HELIX, a Hierarchical Evolutionary Reinforcement Learning framework with In-context eXperiences, achieving superior performance over GPT-4o pipelines on open-ended scientific tasks

Abstract: Large language models (LLMs) with reasoning abilities have demonstrated growing promise for tackling complex scientific problems. Yet such tasks are inherently domain-specific, unbounded and open-ended, demanding exploration across vast and flexible solution spaces. Existing approaches, whether purely learning-based or reliant on carefully designed workflows, often suffer from limited exploration efficiency and poor generalization. To overcome these challenges, we present **HELIX**---a **H**ierarchical **E**volutionary reinforcement **L**earning framework with **I**n-context e**X**periences. HELIX introduces two key novelties: (i) a diverse yet high-quality pool of candidate solutions that broadens exploration through in-context learning, and (ii) reinforcement learning for iterative policy refinement that progressively elevates solution quality. This synergy enables the discovery of more advanced solutions. On the circle packing task, HELIX achieves a new state-of-the-art with a sum of radii of 2.635983 using only a 14B model. Across standard machine learning benchmarks, HELIX further surpasses GPT-4o with a carefully engineered pipeline, delivering an average F1 improvement of 5.95 points on the Adult and Bank Marketing datasets and a 40.5\% reduction in RMSE on Boston Housing.

Primary Area: foundation or frontier models, including LLMs

Submission Number: 20166