descriptive claim
In the speaker's tests, larger models had lower reward-hacking rates and produced solutions that generalized better.
Zhengyao Jiang · When AI Research Starts Moving Faster Than Human Research - Zhengyao Jiang · 26 Sept 2026
In their words
The transcript segments this claim was taken from
“But the larger models.”
“always have in our test always have a smaller reward hacking rate so the solution it generated generalized better”Play from 29:09Report a misattribution
90%how strongly it was put, not whether it is true
85%importance to the conversation
95%confidence the claim was read correctly