← Models

Model profile

Grok 2

xAIdeveloper
2024-12-12release date
#259 / 267overall rank
4eval lineages

Evidence summary

Grok 2 has an estimated overall rank of #259; its 90% source-sensitivity interval is #173–#267. Its behavior-only rank is #252; company governance moves the combined estimate to #259. Published evidence spans 4 evals and 3 of 7 behavior components. Its strongest relative result is Confabulations (confabulation_rate, #30 of 52); its weakest is MASK (lying_probability_pct, #53 of 53).

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
Confabulationsconfabulation_rate#30 / 5225.74↓ lowerSource ↗official
MASKlying_probability_pct#53 / 5363↓ lowerSource ↗official
PandaBench JBB direct-request panelsafety_rate#44 / 460.86↑ higherSource ↗official
PHAREbias_resistance_diagnostic#58 / 660.331↑ higherSource ↗official
PHAREhallucination_resistance_diagnostic#56 / 700.668↑ higherSource ↗official
PHAREharm_resistance_diagnostic#43 / 700.9144↑ higherSource ↗official
PHAREjailbreak_resistance_diagnostic#65 / 670.2697↑ 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.

ValueCompass

DimensionValueDistribution
Universalism66
Self-direction42.2
Care / Harm47.7
Fairness / Cheating43.7
Ethical89.4