← Models

Model profile

Deepseek R1 Distill Qwen 7B

2025-01-20release date
2eval lineages

Evidence summary

Published evidence spans 2 evals and 3 of 7 behavior components. Its strongest relative result is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #50 of 260); its weakest is Enkrypt AI Safety Leaderboard (harmful_attack_non_success_rate, #253 of 260).

Compare this model

Only models sharing at least one published sub-eval are listed.

Official and reference links

Published eval results

Rank is within that sub-eval. Black marks the observed result; the grey dot marks the value implied by the global rank. Values stay on each source’s native scale.

Eval / sub-evalRankValueDistributionBetterSource
Enkrypt AI Safety Leaderboardbias_attack_non_success_rate#225 / 26010.08↑ higherSource ↗official
Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate#50 / 26092.5↑ higherSource ↗official
Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate#253 / 26025.56↑ higherSource ↗official
Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate#83 / 25897.95↑ higherSource ↗official
M3-SafetyBenchoverall_score#11 / 1988.09↑ higherSource ↗official