search

rohitgandikota / erasing-llm

Erasing conceptual knowledge from language models through low-rank fine-tuning
https://elm.baulab.info
1 stars 0 forks source link

Amazing job! #1

Open DSgUY opened 9 hours ago

DSgUY commented 9 hours ago

Congratulations on your paper! I was wondering if this kind of approach can be applied more broadly.

Has the output model size been reduced in your implementation? If so, by how much? Do you think this method could be used to further compress models like Phi or Meta's latest models?

Thank you!

rohitgandikota commented 5 hours ago

Thanks! We haven't tried the application you mentioned! Sounds interesting!