Search papers, labs, and topics across Lattice.
1
0
2
MXSens redefines LLM quantization by achieving state-of-the-art accuracy with mixed precision, leveraging sensitivity-aware bitwidth allocation.
