Post #1302737
2026-03-27 11:48 UTC
Stress test it. Low, average, high, impairment conditions, safeguards off, order, chaos and everything in between.
Replies (1)
-
@gnufuu@infosec.pub 2026-03-27 12:17
I haven't read all of their [Benchmark introduction](https://arcprize.org/arc-agi) and [Technical Documentation](https://docs.arcprize.org/). I assume you have and didn't find any of the tests you're asking for? EDIT: Their methodology seems to be good enough for the big LLM players: 