Working definition
Reproducibility in neuromorphic and biocomputing research requires substrate-specific controls plus a shared provenance record. Silicon studies need device, board, compiler, clock, temperature, and power methods; living systems additionally need donor or line, preparation, maturation, batch, health, contamination, nested replication, and intervention controls.
Mechanism
- Pre-register the unit of analysis and exclusions.
- Version every transformation from substrate to metric.
- Repeat across runs, devices, batches, and sites as appropriate.
Measurements
- Within- and between-unit variance
- Effect size with uncertainty
- Replication and failure rate
Reproducibility controls
- Version hardware, software, firmware, and analysis code.
- Declare dataset, preprocessing, random seeds, and measurement boundary.
- Report repeated runs, variation, exclusions, and failed trials.
Limits and failure modes
- More channels are not more biological replicates.
- Reproducible bias remains bias.
Mathematical connection
Formal structure without substrate erasure
Hierarchical replication evidence
Update effect estimates across runs, devices, biological batches, and laboratories without pooling them as identical units.
Inputs
- Nested observations
- Replication structure
- Prior assumptions
Outputs
- Effect distribution
- Between-unit variation
- Posterior sensitivity
Limit: Hierarchical modeling cannot repair confounded controls, selective reporting, invalid units of analysis, or missing provenance.
Technical and governance sources
- [1]NeuroBench: Advancing Neuromorphic Computing Through Collaborative, Fair and Representative Benchmarking · National Institute of Standards and Technology
Establishes: A community framework separating algorithm and system tracks and defining task, correctness, efficiency, and reporting procedures intended to make neuromorphic results more comparable and reproducible.
Boundary: A benchmark ranks submitted systems on declared tasks and metrics. It does not prove general intelligence, biological equivalence, safety, usefulness outside the benchmark, or superiority under unreported host and data costs.
- [2]Interlaboratory reproducibility of microelectrode array measurements of spontaneous activity in networks of cultured neurons · Journal of Biomolecular Screening
Establishes: An interlaboratory study examining whether specified culture and multielectrode-array procedures can produce comparable spontaneous neural-network measurements across sites.
Boundary: Reproducibility under the studied protocol does not generalize to every cell source, maturation state, array, stimulation regime, analysis pipeline, learning task, or claimed computational capability.
- [3]Guidelines for Stem Cell Research and Clinical Translation · International Society for Stem Cell Research
Establishes: Current professional guidance for oversight, consent, provenance, review, communication, and responsible conduct in stem-cell and organoid research, including research involving human biological materials.
Boundary: Professional guidelines establish governance expectations, not a determination that any organoid is conscious or that all ethical questions are resolved. Local law and independent institutional review still apply.