ModelRefs / AWQ (Activation-Aware Weight Quantization) — AI Glossary

AWQ (Activation-Aware Weight Quantization) — AI Glossary

A post-training quantization method that protects the 1% of salient weights most activated, achieving near-lossless INT4 compression. The counterpart to GPTQ.

Overview

AWQ (Lin et al. 2023) observes that not all weights matter equally—weights corresponding to high-activation channels cause most quantization error. By scaling those channels before quantization, AWQ achieves better INT4 quality than GPTQ with faster on-device inference. Supported by vLLM, TGI, and llama.cpp.

Reference details

Topic	training
Last reviewed	2026-06-24

Commonly confused with

The counterpart to GPTQ. Where GPTQ works from curvature information about the weights themselves, this uses activation statistics to find which weights matter for real inputs, which is why it needs a calibration set drawn from representative data. Both are post-training methods: neither requires retraining, and both are distinct from quantization-aware training, which changes the training run.

Primary source

Lin et al., AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration (arXiv)

Continue your research

Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to AWQ (Activation-Aware Weight Quantization) — AI Glossary.

Frequently asked questions

What is AWQ (Activation-Aware Weight Quantization)?

A post-training quantization method that protects the 1% of salient weights most activated, achieving near-lossless INT4 compression.

What concepts are related to AWQ (Activation-Aware Weight Quantization)?

Closely related concepts include gptq, quantization, gguf.