QMutBench: A Dataset of Quantum Circuit Mutants
Researchers release a massive dataset of over 700,000 faulty quantum circuits to benchmark testing tools.
A team of researchers from institutions including Simula Research Laboratory and the University of Luxembourg has introduced QMutBench, a pioneering dataset designed to address a critical gap in quantum software engineering. The dataset contains a massive collection of over 700,000 quantum circuit mutants, which are intentionally faulty versions of quantum programs. These mutants represent various types of errors, such as incorrect quantum gate operations, that can occur during development. The lack of standardized benchmarks containing known faults has hindered progress in automated quantum software testing, making QMutBench a foundational resource for the field.
QMutBench is accessible through a dedicated online interface that allows quantum software developers and testers to select specific subsets of mutants based on key criteria. Users can filter by the original quantum circuit from which mutants were generated, the desired survival rate (how many mutants a test suite fails to detect), and specific mutation characteristics like the type of faulty gate involved. This enables precise benchmarking for evaluating the effectiveness of new testing techniques or for comparing different approaches. By providing a large-scale, accessible dataset, QMutBench aims to accelerate the development of more robust mutation-guided testing techniques, which are crucial for ensuring the reliability of quantum algorithms as the technology matures.
- Contains over 700,000 quantum circuit mutants representing various fault types.
- Offers an online interface for filtering by circuit origin, survival rate, and mutation characteristics.
- Aims to standardize benchmarking for quantum software testing techniques and tools.
Why It Matters
Provides the first major benchmark for developing reliable testing tools, which is critical for building trustworthy quantum software.