RSS

Chatrik

Karma: 35

Char­ac­ter­iz­ing sta­ble re­gions in the resi­d­ual stream of LLMs

Jett Janiak, jacek, Chatrik, Giorgi Giglemiani, nlpet and StefanHex
26 Sep 2024 13:44 UTC
34 points
3 comments1 min readLW link
(arxiv.org)

Eval­u­at­ing Syn­thetic Ac­ti­va­tions com­posed of SAE La­tents in GPT-2

Giorgi Giglemiani, nlpet, Chatrik, Jett Janiak and StefanHex
25 Sep 2024 20:37 UTC
16 points
0 comments3 min readLW link
(arxiv.org)