Skip to content

Benchmark

Sayf RedSage: cybersecurity_knowledge_frameworks accuracy

Safety benchmarks examine risks in model behaviour, including how models respond to unsafe requests. Higher scores are better.

Models measured
6
Scores recorded
6
Sources
1
IQ weight
0.91

Score distribution

0.856
Mean
0.862
Median
0.03
Std dev
0.83
10th percentile
0.902
90th percentile
0.811
Min
0.8110.902
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.91 for chat models, measured over 6 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: Sayf-eval. Copyright (c) 2026 Qatar Computing Research Institute (QCRI). MIT License. (licence)