logoalt Hacker News

beeringyesterday at 6:30 PM0 repliesview on HN

With Opus it’s hard to tell what was due to the tokenizer changes. Maybe using more tokens for the same prompt means the model effectively thinks more?