icassp-reproducibility
Use when strengthening ICASSP reproducibility across signal-processing modalities — pinning the scoring ruler for the paper's metric, dataset versions and splits, front-end/DSP settings, seeds, and compute, and mapping each claim to a checkable location, since ICASSP has no reviewed appendix and the four pages plus a public release must carry it.
Profile →npx skills add brycewang-stanford/Awesome-Journal-Skills --skill icassp-reproducibility --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# ICASSP Reproducibility Use this before submission and again before camera-ready. ICASSP has **no reviewed supplement**, so reproducibility rests on what the four pages state plus whatever you release publicly (which, under single-blind review, may be public immediately). The recurring ICASSP failure is not a missing repository — it is a **number whose measurement cannot be reconstructed**. ## The evidence spine Map each claim — an algorithm result, a theoretical bound, or an empirical metric — to a checkable location in the paper or the released package: - For an empirical result: dataset and version, split or trial list, front-end/DSP settings, model, the **exact scorer and its configuration**, seeds, number of runs, and reported spread. - For an estimation/detection result: the signal and noise model, the estimator, and the reference bound (e.g., Cramér-Rao) the result is compared against. - For a real-time or embedded claim: hardware, latency or real-time factor, and memory. - Explain any data you cannot release honestly, and describe how a reader could reproduce from the licensed source. ## The scoring ruler is the thing that decays Across ICASSP's modalities, the same trap r
- The evidence spine
- The scoring ruler is the thing that decays
- Front-end determinism
- Degrees of reproducibility
- Reproducibility stub
- Vignette: a keyword-spotting paper
- Output format
Pin the environment and the ruler; regenerate the headline number. pip install -r requirements.txt # exact versions, including the DSP/feature lib python3 run_eval.py --config configs/main.yaml --seed 1 python3 run_eval.py --config configs/main.yaml --seed 2 python3 run_eval.py --config configs/main.yaml --seed 3 python3 aggregate.py --runs runs/ --report mean_std # matches Table 1 mean ± spread
What does the icassp-reproducibility skill do?
Use when strengthening ICASSP reproducibility across signal-processing modalities — pinning the scoring ruler for the paper's metric, dataset versions and splits, front-end/DSP settings, seeds, and compute, and mapping each claim to a checkable location, since ICASSP has no reviewed appendix and the four pages plus a public release must carry it.
How do I install it?
Run `npx skills add brycewang-stanford/Awesome-Journal-Skills --skill icassp-reproducibility --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From brycewang-stanford/Awesome-Journal-Skills, a repository with 909 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.