Loading the model data…
1 · The graph is the selector
—Loading the model data…
root
—
Legend and level
level
0 dropped
2 · Two prompts, one question
ungraded
Loading the model data…
—
Token counts follow the selected vendor: cl100k
counted in this browser, Anthropic’s own count when an Anthropic key is set, and a measured
×1.77 estimate of it when none is. A graded run sends this text plus
one line naming the answer keys, exactly as the benchmark does.
3 · Scoreboard
0 runsNo runs yet