this post was submitted on 18 Mar 2025
9 points (90.9% liked)

Futurology

2512 readers
21 users here now

founded 2 years ago
MODERATORS
[email protected]
[email protected]
[email protected]
[email protected]
9
Claude Sonnet 3.7 (often) knows when it’s in alignment evaluations — Apollo Research (www.apolloresearch.ai)
submitted 1 month ago by [email protected] to c/[email protected]
1 comments fedilink hide all child comments
top 1 comments
sorted by: hot top controversial new old
[–] [email protected] 1 points 1 month ago* (last edited 1 month ago)

This is crazy:

https://images.squarespace-cdn.com/content/v1/6593e7097565990e65c886fd/8389bb0c-1d5f-4f6d-ba91-87ee51504be0/sandbagging_example.png