跳转至

文章背景与核心概要

量子软件开发传统上具有高度的迭代性和易错性，常常受到带噪硬件以及频繁重复执行等复杂因素的阻碍。尽管实验追踪、数据血缘（provenance）和可复现性至关重要，但由于工具链的复杂性和对专业知识的要求，这些实践很难在实际中得到普及。通用大语言模型（LLM）虽然能够辅助开发者，但它们往往会产生幻觉，并且缺乏对领域特定工具的扎实理解（grounding）。

为了解决这些挑战，研究人员推出了 Qlippy，这是一个直接嵌入开发环境中的检索增强生成式AI（GenAI）助手。通过将其响应建立在精心策展的量子软件工程知识库之上，Qlippy 能够上下文化可复现性概念，并将基于 MLflow 的实验追踪（与 QProv 架构保持一致）无缝集成到现有的 Qiskit 程序中。这种解耦架构降低了对模型规模的依赖，为低成本、保护隐私的本地部署铺平了道路。

Qlippy: A Retrieval-Augmented GenAI Assistant for Reproducible Quantum Workflows and Experiment Tracking

Summary

Quantum software development is traditionally iterative, error-prone, and hindered by the complexities of noisy hardware and repeated re-execution. While experiment tracking, provenance, and reproducibility are essential, they are difficult to adopt due to tooling overhead and specialized knowledge requirements. General-purpose Large Language Models (LLMs) can assist developers, but they frequently suffer from hallucinations and lack grounding in domain-specific tools.

To address these challenges, researchers introduce Qlippy, a retrieval-augmented Generative AI (GenAI) assistant embedded directly into the development environment. By grounding its responses in a curated knowledge base of quantum software engineering, Qlippy contextualizes reproducibility concepts and seamlessly integrates MLflow-based experiment tracking (aligned with the QProv schema) into existing Qiskit programs. This decoupled architecture reduces model-scale dependency, paving the way for cost-effective, privacy-preserving, local deployments.

Document Details

Metadata Details
arXiv ID arXiv:2609.05039
Primary Subject Quantum Physics (quant-ph)
Secondary Subjects Artificial Intelligence (cs.AI)
Authors Mahee Gamage, Vlad Stirbu
Submission Date September 4, 2026
Conference/Workshop Accepted for publication in the QGenAI Workshop at IEEE QCE 2026
DOI 10.48550/arXiv.2609.05039

Abstract

Quantum software development is iterative and error-prone. Noisy hardware and repeated re-execution make experiment tracking, provenance, and reproducibility essential, yet these practices are hard to adopt because of tooling complexity and the specialized knowledge they demand. General-purpose language models can help but tend to hallucinate and lack grounding in domain-specific tooling. We present Qlippy, a retrieval-augmented GenAI assistant embedded in the development environment that grounds its responses in a curated corpus of quantum-software-engineering knowledge. Qlippy explains reproducibility and provenance concepts in context and augments existing Qiskit programs with MLflow-based experiment tracking aligned to the QProv schema. By separating knowledge from model parameters, grounding gives explicit control over the scope and provenance of the assistant's responses and reduces reliance on model scale, which points toward low-cost, privacy-preserving local deployment.

Full-Text & External Resources