English

DAQ: Delta-Aware Quantization for Post-Training LLM Weight Compression

Machine Learning 2026-03-25 v1 Artificial Intelligence

Abstract

We introduce Delta-Aware Quantization (DAQ), a data-free post-training quantization framework that preserves the knowledge acquired during post-training. Standard quantization objectives minimize reconstruction error but are agnostic to the base model, allowing quantization noise to disproportionately corrupt the small-magnitude parameter deltas (ΔW\Delta W) that encode post-training behavior -- an effect we analyze through the lens of quantization as implicit regularization. DAQ replaces reconstruction-based objectives with two delta-aware metrics -- Sign Preservation Rate and Cosine Similarity -- that directly optimize for directional fidelity of ΔW\Delta W, requiring only the base and post-trained weight matrices. In a pilot FP8 study, DAQ recovers style-specific capabilities lost under standard quantization while maintaining general performance.

Keywords

Cite

@article{arxiv.2603.22324,
  title  = {DAQ: Delta-Aware Quantization for Post-Training LLM Weight Compression},
  author = {Xiaoming Yu and Shize Tang and Guanghua Yu and Linchuan Xie and Song Liu and Jianchen Zhu and Feng Li},
  journal= {arXiv preprint arXiv:2603.22324},
  year   = {2026}
}
R2 v1 2026-07-01T11:33:52.407Z