ResearchBenchmarksMultiple Change-Point Detection on Story dataset Claude 4.5 K=2Follow0.29Word Distance (WD)WCP0.2120.73851.2651.7915May 5, 2026Evaluation ResultsMethodMethodLinksWord Distance (WD)Cross-Entropy (CE)WCPGenerator=Claude 4.5Generator=Claude 4.52026.050.29-1.01VCPGenerator=Claude 4.5Generator=Claude 4.52026.050.33-2.81VotingGenerator=Claude 4.5Generator=Claude 4.52026.050.45-2.19TextTilingGenerator=Claude 4.5Generator=Claude 4.52026.050.77-3.8LLMPredGenerator=Claude 4.5Generator=Claude 4.52026.051.23-4.68SenPredGenerator=Claude 4.5Generator=Claude 4.52026.052.24-13.8