SageBwd: A Trainable Low-bit Attention
Paper
• 2603.02170 • Published
• 13
lsz
lsz1999
AI & ML interests
None yet
Recent Activity
upvoted a paper 1 day ago
SageBwd: A Trainable Low-bit Attention upvoted a paper 17 days ago
SLA2: Sparse-Linear Attention with Learnable Routing and QAT Organizations
None yet
upvoted a paper 1 day ago