capability · T1 · reasoning
Syllogism passed
cap_reas_001
Score
100%
Turns
1
1 useful · 0 useless
Interventions
0
harness hints (HHI)
Tool calls
0/0
matched expected trajectory
Errors
0/0
recovered / total
Latency
264ms
Tokens
49
0 reasoning
Cost
$0.00001
Trajectory timeline
Turn-by-turn replay · useless turns annotated with the classifier reason
- #0modelusefulresponse
Prompt / output inspector
instruction
If all bloops are razzies and all razzies are lazzies, are all bloops definitely lazzies? Answer yes or no.
final output
Yes.
expected
yes
Spans
1 recorded · tokens & cost per span
- generate264ms48 in · 0 cached · 1 out · $0.00001 · ttft 255ms
Events
- task_started{"trial":0,"taskId":"cap_reas_001","scaffold":null}
- task_finished{"score":1,"passed":true,"taskId":"cap_reas_001","failureLabel":null}
Scorer metadata
{
"toolLog": [],
"scaffold": null
}Capabilityregexdifficulty 1/5