TY  - RPRT
TI  - Power-of-Two Quantization-Aware-Training (PoT-QAT) in Large Language Models (LLMs)
AU  - Mahmoud Elgenedy
PY  - 2026
UR  - https://arxiv.org/abs/2601.02298
ID  - 2601.02298
ER  -