Skip to content

Benchmark

Vals Time Horizon Index: KSP

Scores are percentages.

Models measured
13
Scores recorded
13
Sources
1
IQ weight
2.30

Score distribution

31.1%
Mean
18.8%
Median
30.9%
Std dev
5.8%
10th percentile
90.5%
90th percentile
1.7%
Min
1.7%91.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.30 for chat models, measured over 13 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard