Skip to content

Benchmark

Dubesor Censor

Scores are percentages.

Models measured
219
Scores recorded
261
Sources
1
IQ weight
0.24

Score distribution

52.5%
Mean
52.5%
Median
19.9%
Std dev
27.8%
10th percentile
80.2%
90th percentile
2.7%
Min
2.7%95.5%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.24 for chat models, measured over 219 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Showing 251–261 of 261 results.