Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multiple Choice Questioning (MCQ) on private CCTV 80 videos (test)
Loading...
88.75
Accuracy (MCQ Test)
+ RL
45.85
56.9875
68.125
79.2625
May 21, 2026
Accuracy (MCQ Test)
Updated 12d ago
Evaluation Results
Method
Method
Links
Accuracy (MCQ Test)
+ RL
Training=SFT+RL
2026.05
88.75
+ CCTV SFT
Training=SFT
2026.05
86.25
+ Dashcam SFT
Training=SFT
2026.05
86.25
Gemini 2.5 Pro
Training=0-shot
2026.05
82.5
Gemini 3.1 Flash
Training=0-shot
2026.05
80
CR2
Training=0-shot
2026.05
47.5
Feedback
Search any
task
Search any
task