QuantMLA: Function-Aligned Dual-Path Quantization for Low-Bit MLA KV Caching
Paper • 2609.36760 • Published • 7
Kaneki
Kanashimizzz
AI & ML interests
None yet
Recent Activity
upvoted a paper 3 days ago
QuantMLA: Function-Aligned Dual-Path Quantization for Low-Bit MLA KV Caching updated a collection 2 months ago
Hui updated a collection 2 months ago
HuiOrganizations
None yet