This collection comprises AP-MAE models trained on attention heads from the 3B, 7B, and 15B versions StarCoder2. Currently anonymized for paper review
LaughingLogits
LaughingLogits
AI & ML interests
None yet
Organizations
None yet