Taewhoo Lee

Outlier-Safe Pre-Training for Robust 4-Bit Quantization of Large Language Models

2025 · arXiv (Cornell University)

Extreme activation outliers in Large Language Models (LLMs) critically degrade quantization performance, hindering efficient on-device deployment. While channel-wise operations and adaptive gradient scaling are recognized causes, practical mitigation remains challenging. We introduce Outlier-Safe Pre-Training (OSP), …