Tail-Aware HiFloat4: W4A4 Post-Training Quantization for Wan2.2

Computer Science > Artificial Intelligence [Submitted on 26 May 2026] Tail-Aware HiFloat4: W4A4 Post-Training Quantization for Wan2.2 Zhanfeng Feng, Shuai Guo, Xin Di, Long Peng, Yang Cao, Zhengjun Zha This report describes Tail-Aware HiFloat4, our submission to the low-bit text-to-video generation quantization challenge. Our method adapts the public ViDiT-Q post-training quantization pipeline to Wan2.2 under the HiFloat4 numerical format. We quantize the main linear layers in both Wan2.2 transformer modules with W4A4 HiFloat4 fake quantization, keep numerically sensitive boundary modules in high precision, and introduce an activation-tail-aware percentile calibration module for channel-mask construction. Together with compact PTQ-state restoration, this design reduces the influence of rare calibration outliers while keeping the runtime HiFloat4 arithmetic and sampling pipeline unchanged. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2605.26628 [cs.AI]   (or arXiv:2605.26628v1 [cs.AI] for this version)   https://doi.org/10.48550/arXiv.2605.26628 Focus to learn more Submission history From: ZhanFeng Feng [view email] [v1] Tue, 26 May 2026 07:04:22 UTC (3,993 KB) Access Paper: HTML (experimental) view license Current browse context: cs.AI < prev   |   next > new | recent | 2026-05 Change to browse by: cs References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)