ResearchBenchmarksDetection on Algospeak Paraphrase 1.0 (test)Detection on Algospeak Paraphrase 1.0 (test)Follow0.99Adjusted R2Claude-5.1148-3.5299-1.945-0.3601May 7, 2026Evaluation resultsMethodMethodLinksAdjusted R2Spearman Correlation SignificanceMajority Fit EstimationSignificance CountClaude2026.050.99———Grok2026.050.97———GPT-4o-m2026.050.95———Qwen2026.050.95———GPT-4o2026.050.93———Llama2026.050.28———Mistral2026.05-4.88———Significance Count2026.05———7