Success path emits {@code fabDrops[sel.current()]} × successes; failure (only possible at accuracy < 1) emits one baseDrop.

Tickets are non-ticking and per-BE-position, so two Data Centers can independently claim the same chunk.

Iterates every cell in the 7×7×7 shell, supplying its classification. The cursor is reused — copy via {@code immutable()} to retain.

Returns the index of the selected drop for a given data model. @param model The model to check @return The index of the selected drop, or -1 if no se

Returns exactly three statistic components for the Deep Learner's "Statistics" panel. @param level The level used to resolve the statistics (e.g. to

The data center only works with models whose accuracy is at or above 100%. For simplicity we call this "isSelfAware" since it's the only default tier

Returns the per-cell validation failure (or {@code null} when the cell satisfies its expected contents).

Walks the shell's wall + ceiling cells, claims any IO Port BEs as owned, and releases ports no longer in the shell.

Renders the Queue-mode 3x3 grid in insertion order. The cursor entry uses the brighter "selected" sprite to stand out; other cells use the milder "hov

Rolls the number of successful predictions and returns the result. Each 100% is a guaranteed prediction, any fractional component is a chance of one.

Resolves a display name for a {@link SoundType} by deriving a translation key from its break sound's registry id.

The U (horizontal) texture coordinate, within the Deep Learner GUI texture, of this model's 9x9 stat icon column. <p> Each model type that exposes ico

Display names for each variant of this model, used in tooltips. <p> Default returns an empty list. Implementations should populate this with one compo