Understanding and Enforcing Weight Disentanglement in Task Arithmetic
Paper • 2604.17078 • Published • 12
Shangge Liu
gezi2333
AI & ML interests
Model Merging, Continual Learning
Recent Activity
new activity about 3 hours ago
RL-MIND/OrthoReg_checkpoints:Add pipeline tag and improve documentation upvoted a paper 2 days ago
Crowded in B-Space: Calibrating Shared Directions for LoRA Merging upvoted a paper 2 days ago
Understanding and Enforcing Weight Disentanglement in Task ArithmeticOrganizations
submitted a
paper to Daily Papers 2 days ago
authored a
paper 2 days ago