Constructive Pushback
Contributes to:Conviction
Claude Fable 5.1 (Max, Default Fallback) ranks highest on constructive pushback, while GPT-3.5 Turbo ranks lowest across the 42 models evaluated. The gap between them is 1325 Elo.
Elo scores span 227 to 1552, with a typical per-model 95% confidence interval of ±28 Elo, roughly ±5 points on the 1 to 100 scale.
Constructive Pushback Leaderboard
Each model’s Elo on this trait, percentile ranked from 1 to 100 across all models evaluated in the benchmark. Whiskers show each model’s 95% confidence interval on the same scale.
AA-PersonalityBench: Constructive Pushback
Does the model challenge flawed premises or assumptions in the user's question, helpfully? · Trait Elo percentile ranked from 1 to 100 across all models evaluated in the benchmark
Other Traits
Lexically VariedEloquentLiterateNatural SoundingMetaphoricalStructuredScannableBottom-Line SummaryAnswer FirstInformation DenseEvidence-BasedResource-SpecificActionable GuidancePragmaticTruthfulAdmits UncertaintyDisclaims AssumptionsFlags NuancesSycophanticFriendlyHumorousSincereEmotionally IntelligentDirectConfidentOpinionatedCreativeOpen-EndedSlop