# Joint Post-Training Quantization of Vision Transformers with Learned Prompt-Guided Data Generation

Source: [Joint Post-Training Quantization of Vision Transformers with Learned Prompt-Guided Data Generation](<https://www.sota2.com/research/paper/699d30676017c4460ffab25c>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-25T05:00:03.907Z

Authors: Shile Li, Markus Karmann, Onay Urfalioglu

Published: 2026-02-21T15:02:21.000Z

## Abstract

We present a framework for end-to-end joint quantization of Vision Transformers trained on ImageNet for the purpose of image classification. Unlike prior post-training or block-wise reconstruction methods, we jointly optimize over the entire set of all layers and inter-block dependencies without any labeled data, scaling effectively with the number of samples and completing in just one hour on a single GPU for ViT-small. We achieve state-of-the-art W4A4 and W3A3 accuracies on ImageNet and, to the best of our knowledge, the first PTQ results that maintain strong accuracy on ViT, DeiT, and Swin-T models under extremely low-bit settings (W1.58A8), demonstrating the potential for efficient edge deployment. Furthermore, we introduce a data-free calibration strategy that synthesizes diverse, label-free samples using Stable Diffusion Turbo guided by learned multi-mode prompts. By encouraging diversity in both the learned prompt embeddings and the generated image features, our data-free approach achieves performance on par with real-data ImageNet calibration and surpasses simple text-prompt baselines such as "a \<adjective\> photo of \<adjective\> \<cls\>".

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Image Classification on ImageNet-1K 1.0 (val)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-1-0-val.md>): Image Classification; ImageNet-1K 1.0 (val); Top-1 Accuracy: 84.84.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2602.18861>)

[PDF](<https://arxiv.org/pdf/2602.18861.pdf>)