Skip to content

Benchmark

MLEB - SCALR

Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.

Models measured
18
Scores recorded
18
Sources
1
IQ weight
0.80

Score distribution

68.1%
Mean
67.8%
Median
6.5%
Std dev
61.5%
10th percentile
75.6%
90th percentile
55.2%
Min
55.2%79.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.80 for embedding models, measured over 18 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Data credits: MLEB (Isaacus), MIT (licence)