AI RESEARCH
LoRaQ: Optimized Low Rank Approximation for 4-bit Quantization
arXiv CS.LG
•
ArXi:2604.18117v1 Announce Type: new Post-