Home▸AI & Machine Learning▸Sentencing Disparity Detector (2D)

Sentencing Disparity Detector (2D)

2D statistical simulator: instead of a rotating 3D scatter, watch a severity-vs-sentence plot with a live peer-group corridor, a per-judge disparity bar chart and a z-score histogram all update together as a judge's sentencing bias and the flag threshold change — a simplified model of real judicial-error and sentencing-disparity audits.

AI & Machine Learning2DModerate60 FPS📱 Mobile-adapted⇄ 3D version

AI-assisted judicial-error and bias detection rarely works by reading a verdict and declaring it wrong — it works by comparing one decision against a peer group of similar decisions and flagging the ones that don't fit the pattern. This 2D simulator lays that comparison out as three linked panels: a severity-vs-sentence scatter with a live peer-group corridor, a per-judge disparity bar chart, and a population-wide z-score histogram with the flag threshold marked on it. Dial in a bias for any judge and watch statistically detectable disparity emerge from noise across all three views at once, with live counts of flagged cases and a running disparity index for the judge under review.

⚙ Under the hood

A 2D statistical simulator with three linked panels — a severity-vs-sentence scatter with a live peer-group corridor, a per-judge disparity bar chart, and a population z-score histogram — that flags cases statistically outside their peer group as sentencing bias and the flag threshold change, a simplified model of how real judicial-error and sentencing-disparity audits work.

ai-ethicslegal-techstatisticsbias-detectionanomaly-detectionz-score

2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install