Live intelligence

Latest AI updates

1 updates
Research Editor selected

SignRot: Outlier-Aware Sign-Adjusted Rotation for LLM Quantization

SignRot is a quantization method that addresses massive outlier activations in LLMs by adjusting the signs of Hadamard rows, collapsing bimodal activation distributions into a unimodal profile that is easier to quantize. It is a lightweight, drop-in improvement for Hadamard-based quantization methods, requiring only a single calibration forward pass and adding only element-wise sign flips at inference. Applied to QuaRot on LLaMA3-70B, SignRot achieves a WikiText2 perplexity of 5.88 and an average zero-shot reasoning accuracy of 72.73, surpassing SpinQuant. Code is available on GitHub.

Why it matters: Offers a practical, low-overhead enhancement to LLM quantization with strong reported results on a 70B model, useful for efficient inference work.

LLM quantizationefficient inferenceactivation outliers