logoalt Hacker News

dt3fttoday at 6:47 AM2 repliesview on HN

Opus 4.8:

Which days in a week have the letter d in them?

Response:

Four: Monday, Tuesday, Wednesday, and Sunday.


Replies

abrkntoday at 8:36 AM

I can't reproduce this. Both high and low effort got it right

FrozenSynapsetoday at 6:49 AM

It seems like they’ve been optimising their models for coding. That’s what the benchmarks used in the article suggest at least.