Atlas

Benchmarks

SPU (task-adjusted, refusals excluded)

The headline of UniPat's SPU paper and blog: each configuration's mean Spearman score over the public tasks it was scored on, adjusted for task difficulty by a model fitted jointly across all 13 configurations. Refusals and unrecoverable runs are dropped, so each configuration has its own denominator (30 to 100 tasks). Diagnostic; the primary `spu-v1-public-100` counts every run.

Models
10
Top score
0.5
Leader
Claude Opus 5.5
Updated
Nov 11, 2026
Resolution
—
Saturation
—

No signal or difficulty estimate. Held out of the fit by design.

Organizations
10 dated models · 3 frontier modelsClosed weightsOpen weights
Data: UniPat AIatlas.village.dev

SPU (task-adjusted, refusals excluded) leaderboard

Related benchmarks

Loading Atlas data…