Zhenan Sun

Linear attention's recurrent state can be compressed down to 4 bits while outperforming uniform INT8 once precision is dynamically allocated to long-lived memory tracks and output-sensitive key rows.

By harnessing the power of micro-forensic primitives, this framework achieves unprecedented adaptability in face anti-spoofing, outperforming existing methods in open-world scenarios.

Compressing a reliable large model via quantization yields Small Language Models that are not only more trustworthy but also more adaptable than those trained from scratch.