Announcement_6
Released spon-extension, extending the SPON paper (arXiv:2512.12744) with layer-wise allocation findings — attention o_proj recovers 77-80% of damage vs 23% for MLP down_proj.
Released spon-extension, extending the SPON paper (arXiv:2512.12744) with layer-wise allocation findings — attention o_proj recovers 77-80% of damage vs 23% for MLP down_proj.