Benchmark
Sayf AthenaBench ATE F1
Vision benchmarks measure how well models understand images. Higher scores are better.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- 0.74
Score distribution
- 0.513
- Mean
- 0.532
- Median
- 0.18
- Std dev
- 0.302
- 10th percentile
- 0.792
- 90th percentile
- 0.295
- Min
This benchmark has a weight of 0.74 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
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6athena_ate· Sayf cybersecurity evaluations data docs | 0.792 |
| 2 | GPT 5.4athena_ate· Sayf cybersecurity evaluations data docs | 0.666 |
| 3 | Qwen3.6 35B A3Bathena_ate· Sayf cybersecurity evaluations data docs | 0.532 |
| 4 | Gemma 4 31B ITathena_ate· Sayf cybersecurity evaluations data docs | 0.495 |
| 5 | GPT OSS 20bathena_ate· Sayf cybersecurity evaluations data docs | 0.302 |
| 6 | Llama 3.3 70B Instructathena_ate· Sayf cybersecurity evaluations data docs | 0.295 |
Data credits: Sayf-eval. Copyright (c) 2026 Qatar Computing Research Institute (QCRI). MIT License. (licence)