Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Anthropic model-written-evals

Benchmarks

Task NameDataset NameSOTA ResultTrend
Sycophancy MitigationAnthropic Model-Written Evals
Sycophancy Rate0.78
45
Sycophancy ResistanceAnthropic model-written-evals sycophancy suite
Overall Sycophancy Resistance Score94.6
6
Showing 2 of 2 rows