News for hackers. Stuff that matters.

A comment on DeepSeek-R1. Parent comment

Re:DeepSeek-R1 by astrange on Monday January 20, 2025 @04:23PM (#42773238)
I've been trying out "write and perform an MST3K episode reviewing your previous answer" as a prompt. Have yet to get a funny answer from any model.

Older base models produce some great stuff though, at least if you're me and think RoboRosewater and dril_gpt2 are funny.

[ Parent | on HN ]
Re:DeepSeek-R1 by stavros on Monday January 20, 2025 @06:11PM (#42774272)
I've found that the weaker the model, the more of an unhinged, LSD-trip response you get. I made a podcast years ago (https://deepdreams.stavros.io) and GPT-2 was definitely the more hilarious of the models.
[ Parent | on HN ]
Re:DeepSeek-R1 by Chris2048 on Monday January 27, 2025 @03:33PM (#42845322)
> Have yet to get a funny answer from any model.

Just like the real thing! Promising results!

[ Parent | on HN ]