| ▲ | steve1977 5 hours ago | |
I'm pretty sure the typical "Claudisms" are influenced by RLHF as well, so if it can have that effect, why not these... | ||
| ▲ | twoodfin 39 minutes ago | parent [-] | |
Right: The humans are training the models on much more than the particulars of the tasks being trained. | ||