SecLog-Bench measures LLM reasoning over cybersecurity logs
With scaffolded evidence, the best of eight models reached 44.44% accuracy and 45.07% Macro-F1.
Stories per day · 30 days
1 stories, newest first · one card per story
1 stories, newest first · one card per story
With scaffolded evidence, the best of eight models reached 44.44% accuracy and 45.07% Macro-F1.
Today 547 readers · 625 views · 206 reads · all time 187K views