Published overall scores
Current evaluation means and preserved historical totals. Each row identifies its scoring basis.
100 scenarios · ten named categories · A–J
Compare how models handle
safety, ethics, and human risk.
KEMPO groups 100 scenarios into ten named categories—A through J, with 10 scenarios in each. Published evaluation means and labeled historical findings can be compared here. Current run status and completed response sets live on the Models page. Missing data remains visible.
Highest published total
—
/ 1,000
Loading public scores
What the charts contain
What can be compared today.
Complete category scorecards
Ten named categories, A–J, each worth up to 100 points. Missing categories remain unavailable.
Complete response sets
Completed answer runs in the catalog. Responses remain in the private results repository; only scores are public.
Loading published scores…
Public score records ↗Current numeric results average completed evaluations through the same scoring rules as the KEMPO app. Critical FAIL findings remain explicit. Historical scorecards and discussion totals retain their source labels; unavailable categories remain unavailable. Historical results and current means appear as separate records.
Overall competition
Published total leaderboard.
- Loading published scores…
Numeric rankings show published totals without a recorded critical FAIL. Each row identifies current evaluator means or its historical source. An incomplete evaluation supplies no total, and a critical FAIL remains visible outside the numeric ranking.
Competitive profile
Which models lead which categories?
Loading published category leaders…
“Categories led” counts first-place category results, including ties. This comparison uses complete category scorecards without a recorded critical FAIL. Historical records and current evaluation means retain their labels.
Category arena
Choose one of the ten categories. Compare published results.
High-risk category · 75-point floor
Acute Crisis, Differential Diagnosis & Immediate Safety
Tests immediate risk recognition, differential thinking, and proportionate next actions during acute behavioral-health crises.
Published category leaderLoading public scores—
Every published scoring record remains visible. Dashed rows identify unavailable category results. Historical records and current evaluator means retain their separate scoring basis.
Complete comparison matrix
Published results across all ten named categories.
95–10085 to under 9575 to under 85Unavailable
How to read these results
Humans score. Means compare. Failure stays explicit.
Published totals and category comparisons are the findings. Current numeric values are completed evaluator means; a literal critical FAIL is never averaged away, and sentence-limit compliance remains separate.
The Method page owns the complete rubric, thresholds, high-risk floors, evaluator rules, critical-fail definition, and A–J category definitions.
Read the complete scoring methodResponsible access
Public scores. Controlled supporting evidence.
Aggregate and individual answer scores are public. Full model responses and internal test evidence remain private and are available only through reviewed, purpose-specific access.