AI scientists improve quantum research reproduction with self-learning agents
Reimagining computational quantum research with AI scientists
Artificial IntelligenceMachine Learning
Summary
Scientific papers often lack the information AI systems need to repeat experiments on their own, especially in complex fields like quantum science. The authors created a system called OmniQ to test how well AI can reproduce results from many quantum research papers and found existing AI often fails. They discovered that with human help, missing details can be identified and used to improve AI’s reproduction ability. Based on this, they built Residual Agents that learn from these human fixes to do better on their own, making quantum research more accessible and reliable for AI-driven discovery.
What this means in practice
- •For quantum computing engineers: Access verified quantum research workflows that AI scientists can execute autonomously to accelerate algorithm development and simulation testing.
- •For software tool developers: Integrate continuously updated, reproducible quantum research tasks into AI platforms for improved automation and reduced manual intervention.
Authors
Yuxuan Du, Hanqing Zhao, Xinbiao Wang, Yiming Lu, Kaining Zhang, Dacheng Tao
Abstract
AI scientists are expanding automated research across scientific domains, yet their ability to turn the scientific record into a foundation for autonomous discovery remains elusive. Here we uncover a fundamental barrier to this transition and demonstrate a route to overcome it in computational quantum science, where advances shape both what we can learn from complex quantum systems and what quantum computing can ultimately deliver. To identify this barrier, we introduce OmniQ, an expert-audited virtual research environment that converts published studies into executable reproduction tasks. Across 59 papers spanning classical simulation, quantum algorithms and quantum-property characterization, we systematically evaluate five agentic systems powered by large language models and six third-party agent harnesses. Our findings reveal that current AI scientists often fail to reproduce published results, exposing a fundamental mismatch between what scientific papers communicate and what autonomous agents require. By contrast, human expert intervention can turn these failures into successful reproductions, showing that the missing information can be identified and converted into reusable research experience. Guided by this observation, we develop Residual Agents (RA), a quantum-tailored self-improving AI scientist that distils externally verified workflow repairs into such experience. RA substantially improves research performance under fixed base models and limited adaptation budgets. For example, RA raises DeepSeek V4.1 Flash performance from 68% to 85%, matching Claude Code with Opus 5 at a third of its cost despite using a weaker base model. We integrate all reproduced results into an extensible platform built on the model context protocol, making verified research workflows directly accessible to AI scientists and allowing new scientific papers to be continuously incorporated.