Rethinking Cross-Layer Information Routing in Diffusion Transformers
Paper • 2605.20708 • Published • 111
qiubo
thenext
·
AI & ML interests
None yet
Recent Activity
upvoted a paper about 1 month ago
Rethinking Cross-Layer Information Routing in Diffusion Transformers upvoted a paper about 1 month ago
Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps upvoted a paper almost 2 years ago
Inference Performance Optimization for Large Language Models on CPUsOrganizations
