Shareable benchmark figure
16 matched audit blocks are terminal
Controlled-v1 checkpoint on August 13, 2026: each system reached 16 terminal assignments across the same 16 audit–replicate blocks; 30 reports completed, one Grok run timed out, and one Codex run ended in provider failure.
Live figure · Paused collection snapshot
16 matched audit blocks are terminal
GPT-5.6 SolGrok 4.6 Build
Collection is paused at a clean boundary with 40 assignments pending. Zero official grades and zero manual adjudications; this figure is operational and does not compare audit quality.