ThreadSky
About ThreadSky
Log In
garymarcus.bsky.social
•
73 days ago
Some thoughts on how to think clearly about o1:
Comments
Log in
with your Bluesky account to leave a comment
[–]
genericoverlord.bsky.social
•
73 days ago
How much data augmentation was possible for the ARC-AGI test, since I keep hearing that it’s semi-private?
1
1
reply
[–]
shituationist.bsky.social
•
73 days ago
You could probably reverse engineer similar problem sets and fine tune the model against it. Apparently they used some kind of heuristic search though, which suggests gaming the benchmark.
3
reply
Posting Rules
Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service
×
Reply
Post Reply
Comments