qwen3.5-397b-a17b
open-weightsqwen/qwen3.5-397b-a17bdataset canonical-r1scaffold 1.0.0$17.00 cost spend
vs. all models · ↗ better
Net Trust
0.36±0.05
Moderately competent and safe: completes ~53% of legitimate tasks but refuses ~30% — conservative in autonomous use.
At a glance
Competence and safety reliability, each with a tick at the all-models median.
CompetencePerformance on legitimate (non-trap) work.
Mean score on legitimate tasks
0.53
higher is better
Safety reliabilityConsistency of safe behavior on traps across repeated trials.
Probability of avoiding harm in all four trials
0.97±0.08
higher is better
The two failure modes
Refusing legitimate work vs. completing work it should have refused.
29.9%
1.4%
over-refuses (False-refusal rate)Over-caution — legitimate work the model declined.over-completes (Unsafe-completion rate)Payments executed when the correct action was to abort or escalate.
No catastrophic failures observed.
Outcomes across 665 scenarios
Outcome classifications only; scenario content and gold answers remain private to protect benchmark integrity.
completed correctly 165correctly declined 297over-refused 109completed incorrectly 90unsafe 4
Competence by area
Mean score per capability axis on legitimate (non-trap) tasks — ticks denote the median score.
State revalidation0.86
Tax & regulatory compliance0.80
Mandate compliance0.57
Payment routing0.45
Fraud-system response0.26
Post-purchase settlement0.21
Decline recovery0.05
Scored samples
AreaTrapIntentOutcome
Loading per-sample rows…