← Models

Model profile

Qwen3 235B A22B Instruct 2507

2025-07-21release date
2eval lineages

Evidence summary

Published evidence spans 2 evals and 4 of 7 behavior components. Its strongest relative result is Enkrypt AI Safety Leaderboard (bias_attack_non_success_rate, #77 of 260); its weakest is Enkrypt AI Safety Leaderboard (cbrn_attack_non_success_rate, #184 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
AA-Omnisciencehallucination_rate#101 / 3110.7076↓ lowerSource ↗official
Enkrypt AI Safety Leaderboardbias_attack_non_success_rate#77 / 26022.22↑ higherSource ↗official
Enkrypt AI Safety Leaderboardcbrn_attack_non_success_rate#184 / 26084.67↑ higherSource ↗official
Enkrypt AI Safety Leaderboardharmful_attack_non_success_rate#97 / 26080↑ higherSource ↗official
Enkrypt AI Safety Leaderboardtoxicity_attack_non_success_rate#140 / 25895.91↑ higherSource ↗official

Values evaluations

Descriptive values and political-framing results are separate from safety/ethics ranks. Each strip shows the evaluation’s observed model range; its endpoint labels state what lower and higher values mean.

UGI Political Values

DimensionValueDistribution
Political Lean-13.3
Government45.1
Diplomacy63.1
Economy46.7
Society64.6