Outlier-Safe Pre-Training for Robust 4-Bit Quantization of Large Language Models Paper • 2506.19697 • Published Jun 24, 2025 • 45
