# P4Q: Learning to Prompt for Quantization in Visual-language Models

arXiv: https://arxiv.org/abs/2409.17634 · HTML: https://mlsanity.com/paper/2409.17634/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Placed in Visual token pruning and efficiency, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 27 of its 34 references that are in the map.
- First posted on arXiv 26 Sep 2024.

## Placed in

- Multimodal LLMs / vision-language models / [Visual token pruning and efficiency](https://mlsanity.com/mllm/visual-token-pruning-and-efficiency/)

## Builds on (most cited first, 27 of 27)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [Learning to Prompt for Vision-Language Models](https://mlsanity.com/paper/2109.01134/) (2021)
- [The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization](https://mlsanity.com/paper/2006.16241/) (2020)
- [FitNets: Hints for Thin Deep Nets](https://mlsanity.com/paper/1412.6550/) (2014)
- [Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference](https://mlsanity.com/paper/1712.05877/) (2017)
- [Do ImageNet Classifiers Generalize to ImageNet?](https://mlsanity.com/paper/1902.10811/) (2019)
- [Conditional Prompt Learning for Vision-Language Models](https://mlsanity.com/paper/2203.05557/) (2022)
- [Natural Adversarial Examples](https://mlsanity.com/paper/1907.07174/) (2019)
- [Learning Robust Global Representations by Penalizing Local Predictive Power](https://mlsanity.com/paper/1905.13549/) (2019)
- [Open-vocabulary Object Detection via Vision and Language Knowledge Distillation](https://mlsanity.com/paper/2104.13921/) (2021)
- [Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model](https://mlsanity.com/paper/2203.14940/) (2022)
- [Factual Probing Is [MASK]: Learning vs. Learning to Recall](https://mlsanity.com/paper/2104.05240/) (2021)
- [Predicting Parameters in Deep Learning](https://mlsanity.com/paper/1306.0543/) (2013)
- [Low-bit Quantization of Neural Networks for Efficient Inference](https://mlsanity.com/paper/1902.06822/) (2019)
- [Forward and Backward Information Retention for Accurate Binary Neural Networks](https://mlsanity.com/paper/1909.10788/) (2019)
- [FQ-ViT: Post-Training Quantization for Fully Quantized Vision Transformer](https://mlsanity.com/paper/2111.13824/) (2021)
- [Training Binary Neural Networks with Real-to-Binary Convolutions](https://mlsanity.com/paper/2003.11535/) (2020)
- [Q-ViT: Accurate and Fully Quantized Low-bit Vision Transformer](https://mlsanity.com/paper/2210.06707/) (2022)
- [Towards Effective Low-bitwidth Convolutional Neural Networks](https://mlsanity.com/paper/1711.00205/) (2017)
- [Q-DETR: An Efficient Low-Bit Quantized Detection Transformer](https://mlsanity.com/paper/2304.00253/) (2023)
- [Recurrent Bilinear Optimization for Binary Neural Networks](https://mlsanity.com/paper/2209.01542/) (2022)
- [Few-Shot Learning with Visual Distribution Calibration and Cross-Modal Distribution Alignment](https://mlsanity.com/paper/2305.11439/) (2023)

## Built on by (most cited first, 7 of 7)

- [A Survey on Efficient Vision-Language Models](https://mlsanity.com/paper/2504.09724/) (2025)
- [Gated Relational Alignment via Confidence-based Distillation for Efficient VLMs](https://mlsanity.com/paper/2601.22709/) (2026)
- [Semantic Alignment and Reinforcement for Data-Free Quantization of Vision Transformers](https://mlsanity.com/paper/2412.16553/) (2024)
- [Harnessing Input-Adaptive Inference for Efficient VLN](https://mlsanity.com/paper/2508.09262/) (2025)
- [D4C: Data-Free Quantization for Contrastive Language-Image Pre-training Models](https://mlsanity.com/paper/2511.15411/) (2025)
- [Arbitrary Ratio Feature Compression via Next Token Prediction](https://mlsanity.com/paper/2602.11494/) (2026)
- [MiX: Micro-Inverted-Scaling for End-to-End Low-Bit Vision-Language Model Acceleration](https://mlsanity.com/paper/2609.19683/) (2026)

## Abstract (arXiv)

Large-scale pre-trained Vision-Language Models (VLMs) have gained prominence in various visual and multimodal tasks, yet the deployment of VLMs on downstream application platforms remains challenging due to their prohibitive requirements of training samples and computing resources. Fine-tuning and quantization of VLMs can substantially reduce the sample and computation costs, which are in urgent need. There are two prevailing paradigms in quantization, Quantization-Aware Training (QAT) can effectively quantize large-scale VLMs but incur a huge training cost, while low-bit Post-Training Quantization (PTQ) suffers from a notable performance drop. We propose a method that balances fine-tuning and quantization named ``Prompt for Quantization'' (P4Q), in which we design a lightweight architecture to leverage contrastive loss supervision to enhance the recognition performance of a PTQ model. Our method can effectively reduce the gap between image features and text features caused by low-bit quantization, based on learnable prompts to reorganize textual representations and a low-bit adapter to realign the distributions of image and text features. We also introduce a distillation loss based on cosine similarity predictions to distill the quantized model using a full-precision teacher. Extensive experimental results demonstrate that our P4Q method outperforms prior arts, even achieving comparable results to its full-precision counterparts. For instance, our 8-bit P4Q can theoretically compress the CLIP-ViT/B-32 by 4 × while achieving 66.94% Top-1 accuracy, outperforming the learnable prompt fine-tuned full-precision model by 2.24% with negligible additional parameters on the ImageNet dataset.