logoalt Hacker News

irthomasthomas • yesterday at 5:44 PM • 1 reply • view on HN

Why does anthropic change the set of benchmarks they use with every new model release?

https://www.anthropic.com/news/claude-opus-4-7

https://www.anthropic.com/news/claude-opus-4-6

Replies

pietz • yesterday at 5:55 PM

1. Benchmarks saturate 2. They select the most impressive improvments