Skip to content
View sparkainlp-x's full-sized avatar
🎯
Focusing
🎯
Focusing
0 followers · 2 following

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Achievements

Achievement: Pull Sharkx2Achievement: YOLO

Block or report sparkainlp-x

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
sparkainlp-x/README.md

Spark AI NLP banner: glowing brain-circuit logo with the words SPARK AI NLP over a dark lab scene

Spark AI NLP

Disability-led applied AI and research software from Fredericton, New Brunswick. I build reproducible telemetry benchmarks, C++/HLS prototypes, source-grounded RAG guardrails, and offline evidence-audit tools.

sparkainlpx.xyz · LinkedIn · ORCID iD 0009-0000-9778-5374 · Zenodo community: Spark AI NLP: Research Software
Founder: Jean-François Brisson · Français (langue maternelle) · English (fluent)

Open to work: applied AI research and research-software roles, remote within Canada or on site in Fredericton. See Contact.

The public repositories are research software and prototypes. They are not hardware, field, or medical products, and they do not report results on quantum hardware.

Start here

  • OES-Resilience — Reproducible telemetry-anomaly benchmark with detector comparisons and a preregistered NASA SMAP/MSL evaluation. Outcome: OES32 did not meet its stated success criterion on that evaluation; see the README for scope and results.

  • OES32-HLS — C++/HLS research prototype with a streaming triage kernel, Python bindings, and bit-exact software tests. Status: software tests run; FPGA synthesis and hardware execution are UNRUN.

  • Evidence Passport — Offline tool that converts an experiment-run manifest into a static HTML evidence passport and normalized JSON. It records declared evidence; it does not independently validate or re-score the experiment.

Evidence labels

Project READMEs use these labels where they report evidence. Check each README for the details behind a REPORTED result.

  • SYNTHETIC: generated inputs or simulation.
  • REPORTED: measured by the project author; not independently reproduced.
  • TARGET: a design goal; not yet measured or achieved.
  • UNRUN: a planned or available run that has not been performed.

Code quality or reproducibility alone does not establish scientific validity.

Projects

Versions are the latest GitHub releases. DOIs are Zenodo concept DOIs (all versions).

Telemetry benchmarks and detectors

  • oes-resilience · v0.6.0: open, reproducible benchmark for 512-channel telemetry anomaly detection with the transparent OES32 reference detector, a stress suite, replay evaluation and a detector plugin API. Baseline results are SYNTHETIC. Release v0.5.0 added a preregistered evaluation on one public dataset (NASA SMAP/MSL) under one locked protocol; OES32 did not meet its stated success criterion there. In v0.6.0, OES-32-v2 (locked protocol, test run once) did not meet its predeclared improvement criterion: test event F1 0.213 SMAP / 0.434 MSL vs 0.379 / 0.435 for v1. DOI 10.5281/zenodo.23071166 (v0.6.0: 10.5281/zenodo.23336253)
  • oes512q-latch · v0.2.1: classical OES-32/OES-512 residual latch with a self-calibrated threshold, compared with standard unsupervised detectors on all 47 labelled series of the public Numenta Anomaly Benchmark (NAB). Mixed results, reported as measured: best block ranking by average precision, behind CUSUM on F1. Not a quantum code. DOI 10.5281/zenodo.22998570
  • multi-quantum-oes · v0.1.2: offline, standard-library OES-512 replay-triage workbench with a preregistered stress evaluation whose result is negative and published as such, plus an isolated "AI ∩ quantum" toy lab. SYNTHETIC data; exact classical simulation, not a QPU. DOI 10.5281/zenodo.23113851
  • oes-telemetry-bench: offline replay bench for synchronized 32-channel OES32 telemetry frames, with a strict input contract and frozen thresholds; synthetic demo. DOI 10.5281/zenodo.23175492
  • spark-membrane: fail-closed console that audits SYNTHETIC 32-channel frames against the OES repositories, pinned by commit SHA, with a hash-chained run trail, evidence passport and claims gate; thresholds uncalibrated; demo page. DOI 10.5281/zenodo.23175490
  • homological-cascade-engine: classical, synthetic signal-processing framework for chain-complex diffusion, Hodge residual scoring and streaming POT thresholds; not a QKD system or adaptive-optics hardware. Synthetic tests do not establish real-telemetry performance or operational false-alarm control. DOI 10.5281/zenodo.23307846
  • oes512-telemetry-engine: asynchronous Python prototype for 512-channel telemetry ingestion, DSP, and monitoring; see its README for current implementation scope. DOI 10.5281/zenodo.23307953
  • oes32-dashboard: PyQt6/PyQtGraph desktop dashboard using synthetic stimuli; not a medical or field system. DOI 10.5281/zenodo.23307955
  • flip-observer-1024: calibration prototype for a classical weighted readout under a hidden channel-order reversal (512 inputs, two noisy banks); SYNTHETIC; demo page. DOI 10.5281/zenodo.23173449
  • pilottrace: offline validation and block scoring of 512-value numeric trace frames. DOI 10.5281/zenodo.23187278

OES-32 contract and implementations

Evidence, provenance and claims audits

  • evidence-passport · v0.1.1: offline standard-library audit tool that turns one experiment-run manifest into a static HTML evidence passport and normalized JSON (fail-closed SHA-256); it records declared evidence and never re-scores it. The bundled OES-Resilience SMAP/MSL sample keeps the verdict "criterion not met"; three further examples are labelled SYNTHETIC; sample pages. DOI 10.5281/zenodo.23165143 (v0.1.1: 10.5281/zenodo.23336285)
  • quantum-claims-passport · v0.1.0: offline claims audit for scientific communication. It classifies public claims (a planned Swiss quantum-computer hub; a modeled 613 THz tubulin frequency band; a rat behavioral study) as announcement, computational model, animal behavioral result, hypothesis, or unsupported inference, with exact-SI conversions and no combined score. Source audit only, snapshot 2026-10-05; it makes none of the audited claims and has no affiliation with the institutions named; report. DOI 10.5281/zenodo.23167801
  • measurement-trail: SHA-256 hash-chained measurement provenance trails. DOI 10.5281/zenodo.23067465

Retrieval-augmented generation

Classical simulations and testbenches

None of these are quantum-hardware results.

Offline data tools

Demos and educational prototypes

Repository documentation varies; check each README for its scope, limits, evidence status and setup.

How I work

  • Contracts first. Each definition lives in one normative place with executable tests. For OES-32 that is oes32-residual@b77b612 (ADR-001); other implementations are labelled Profile A sidecars and document how they differ.
  • Fail-closed where specified. Selected READMEs label unrun or unmeasured work, and results are published with their caveats, including negative outcomes.
  • Reproducibility varies by project. Some repositories document tested quickstarts, automated tests or CI; check the individual README and workflow.
  • Accessibility first. Ask for alternative formats by opening an issue in any repository.

Contribution guidelines, the code of conduct and the security policy are shared across all repositories: sparkainlp-x/.github.

Cite

Use GitHub's "Cite this repository" button or the CITATION.cff and DOI links in each repository. Records are being added to the Spark AI NLP: Research Software Zenodo community.

Contact

Open to applied AI research and research-software roles (remote within Canada, or Fredericton, NB), and to collaboration with research groups, FPGA/decoder engineers, and accessibility-focused partners.

Popular repositories Loading

  1. digital-to-wave-testbench digital-to-wave-testbench Public

    Synthetic, dimensionless testbench for the signed-amplitude sine encoding in M. S. Lee's 'Digital to Wave' preprint: seeded noise/phase/CFO/timing benchmarks, Wilson CIs, hash-locked modules. No ph…

    Python 2

  2. quantum-error-correction-demo quantum-error-correction-demo Public

    Classical Python toy simulation of OES-32-style metrics (no qubits, no QEC code). All outputs SYNTHETIC.

    Python 1

  3. oes32-residual oes32-residual Public

    Normative OES-32 residual reference (ADR-001): deterministic max-abs residual over two 32-vectors, strict > tolerance failure rule, contract tests. SYNTHETIC inputs only.

    Python 1

  4. qldpc_decoder_cpp qldpc_decoder_cpp Public

    C++/HLS qLDPC decoder scaffold: CMake + Catch2 + GF(2) micro-benchmark, Vitis HLS/Vivado/PetaLinux scripts for ZCU111. BP kernel and HIL verification are placeholders; hardware results UNRUN.

    C++ 1

  5. oes32-hls oes32-hls Public

    C++ HLS research prototype of OES-32 telemetry-triage checks for AMD ZCU111, with g++ testbenches and bit-exact pybind11 Python bindings. Synthetic stimuli only; FPGA synthesis UNRUN.

    Jupyter Notebook 1

  6. spark-rag-guardrail spark-rag-guardrail Public

    Source-grounded RAG (ChromaDB + Ollama) with a retrieval-relevance guardrail that refuses instead of guessing

    Python 1