Skip to content

Benchmark

SuperCLUE Total

Scores are percentages.

Models measured
20
Scores recorded
20
Sources
1
IQ weight
2.02

Score distribution

48.7%
Mean
43.2%
Median
11.6%
Std dev
36.4%
10th percentile
61.6%
90th percentile
34.6%
Min
34.6%73.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.02 for chat models, measured over 20 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard