Good work!
#1
by
JackCloudman
- opened
Hey :D Thanks for sharing this abliterated version of the reasoning model. I’m really curious about the process you used—did you do anything specific to achieve these results? When I tried “abliterating” the reasoning section in my own tests, the model ended up refusing to respond. Could you share a notebook or any resources explaining how you managed this? Thanks a bunch!
I second this I would also like to know the "Recipe" for creating this model!