Comment by walrus01

4 days ago

The answer is "it depends", here's GLM5.3 when asked about Tienanmen Square in 1989:

https://ibb.co/gLgFSV0J

Yup, it varies between runs (depending on the seed, most likely), but since the knowledge is here it wouldn't be too hard to nudge the model in the right direction with grpo alone.