Plaice Maturity-at-Age Explorer

Female American plaice, NAFO Divisions 3L, 3N, 3O — from the neural network mixed-effects model in arXiv:2608.31133. The chat panel below is grounded in an interactive walkthrough of the paper via retrieval, for plain-language questions about the model itself.

Ask about this result

Answers cite the confidence interval for anything they quote from the chart.

e.g. “Why is this cohort maturing later?” or “How confident is the model at age 6?”

Training data

Loading…

Limitations & extrapolation risk

The model does not extrapolate outside – or age – — the API rejects requests outside that window rather than guessing. Within the window, the earliest cohorts (roughly 1958–1964, before the survey observed those fish at young ages) and a scatter of the most recent cohorts have a non-identifiable delta-method confidence interval, because the data needed to pin down that specific cohort's deviation doesn't exist. Those cells are served with an honest [0%, 100%] interval rather than a falsely precise one.

Uncertainty flag (current view)

Run a prediction to see uncertainty for the selected range.

Data provenance

Fitted model and training data: arXiv:2608.31133 (Zheng, Cheung, Sharma, Thorson & Cadigan, 2026). Predictions served from a static grid baked from the fitted TMB model (see plaice-api/model/precompute.R) — this app performs lookups against that grid, not live re-fitting.