{"ID":2833446,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.03673","arxiv_id":"2512.03673","title":"ConvRot: Rotation-Based Plug-and-Play 4-bit Quantization for Diffusion Transformers","abstract":"Diffusion transformers have demonstrated strong capabilities in generating high-quality images. However, as model size increases, the growing memory footprint and inference latency pose significant challenges for practical deployment. Recent studies in large language models (LLMs) show that rotation-based techniques can smooth outliers and enable 4-bit quantization, but these approaches often incur substantial overhead and struggle with row-wise outliers in diffusion transformers. To address these challenges, we propose ConvRot, a group-wise rotation-based quantization method that leverages regular Hadamard transform (RHT) to suppress both row-wise and column-wise outliers while reducing complexity from quadratic to linear. Building on this, we design ConvLinear4bit, a plug-and-play module that integrates rotation, quantization, GEMM, and dequantization, enabling W4A4 inference without retraining and preserving visual quality. Experiments on FLUX.1-dev demonstrate a 2.26$\\times$ speedup and 4.05$\\times$ memory reduction while maintaining image fidelity. To our knowledge, this is the first application of rotation-based quantization for plug-and-play W4A4 inference in diffusion transformers.","short_abstract":"Diffusion transformers have demonstrated strong capabilities in generating high-quality images. However, as model size increases, the growing memory footprint and inference latency pose significant challenges for practical deployment. Recent studies in large language models (LLMs) show that rotation-based techniques ca...","url_abs":"https://arxiv.org/abs/2512.03673","url_pdf":"https://arxiv.org/pdf/2512.03673v1","authors":"[\"Feice Huang\",\"Zuliang Han\",\"Xing Zhou\",\"Yihuang Chen\",\"Lifei Zhu\",\"Haoqian Wang\"]","published":"2025-12-03T11:02:16Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Diffusion Model\",\"Transformer\",\"Large Language Model\",\"Language Model\"]","has_code":false}