# Synthetic Data is an Elegant GIFT for Continual Vision-Language Models

arXiv: https://arxiv.org/abs/2503.04229 · HTML: https://mlsanity.com/paper/2503.04229/

## Key facts

- Cited by 21 papers in the map (data through 14 Oct 2026).
- Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 55 of its 82 references that are in the map.
- First posted on arXiv 6 Mar 2025.

## Placed in

- Multimodal LLMs / vision-language models / [Prompt learning and adaptation of CLIP](https://mlsanity.com/mllm/prompt-learning-and-adaptation-of-clip/)
- Synthetic data and self-improvement for LLMs / [Self-improving MLLMs and synthetic data](https://mlsanity.com/synth/self-improving-mllms-and-synthetic-data/)

## Builds on (most cited first, 30 of 55)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Classifier-Free Diffusion Guidance](https://mlsanity.com/paper/2207.12598/) (2022)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [Overcoming catastrophic forgetting in neural networks](https://mlsanity.com/paper/1612.00796/) (2016)
- [Zero-Shot Text-to-Image Generation](https://mlsanity.com/paper/2102.12092/) (2021)
- [Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision](https://mlsanity.com/paper/2102.05918/) (2021)
- [Learning without Forgetting](https://mlsanity.com/paper/1606.09282/) (2016)
- [CLIPScore: A Reference-free Evaluation Metric for Image Captioning](https://mlsanity.com/paper/2104.08718/) (2021)
- [iCaRL: Incremental Classifier and Representation Learning](https://mlsanity.com/paper/1611.07725/) (2016)
- [Learning to Prompt for Vision-Language Models](https://mlsanity.com/paper/2109.01134/) (2021)
- [Describing Textures in the Wild](https://mlsanity.com/paper/1311.3618/) (2013)
- [Fine-Grained Visual Classification of Aircraft](https://mlsanity.com/paper/1306.5151/) (2013)
- [Visual Prompt Tuning](https://mlsanity.com/paper/2203.12119/) (2022)
- [EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification](https://mlsanity.com/paper/1709.00029/) (2017)
- [Conditional Prompt Learning for Vision-Language Models](https://mlsanity.com/paper/2203.05557/) (2022)
- [Continual Learning with Deep Generative Replay](https://mlsanity.com/paper/1705.08690/) (2017)
- [Memory Aware Synapses: Learning what (not) to forget](https://mlsanity.com/paper/1711.09601/) (2017)
- [CLIP-Adapter: Better Vision-Language Models with Feature Adapters](https://mlsanity.com/paper/2110.04544/) (2021)
- [When Does Label Smoothing Help?](https://mlsanity.com/paper/1906.02629/) (2019)
- [Learning to Prompt for Continual Learning](https://mlsanity.com/paper/2112.08654/) (2021)
- [Dark Experience for General Continual Learning: a Strong, Simple Baseline](https://mlsanity.com/paper/2004.07211/) (2020)
- [Robust fine-tuning of zero-shot models](https://mlsanity.com/paper/2109.01903/) (2021)
- [Large Scale Incremental Learning](https://mlsanity.com/paper/1905.13260/) (2019)
- [Lifelong Learning with Dynamically Expandable Networks](https://mlsanity.com/paper/1708.01547/) (2017)
- [DualPrompt: Complementary Prompting for Rehearsal-free Continual Learning](https://mlsanity.com/paper/2204.04799/) (2022)

## Built on by (most cited first, 21 of 21)

- [Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting](https://mlsanity.com/paper/2508.04227/) (2025)
- [Learning to Pose Problems: Reasoning-Driven and Solver-Adaptive Data Synthesis](https://mlsanity.com/paper/2511.09907/) (2025)
- [IAP: Improving Continual Learning of Vision-Language Models via Instance-Aware Prompting](https://mlsanity.com/paper/2503.20612/) (2025)
- [Spectral Imbalance Causes Forgetting in Low-Rank Continual Adaptation](https://mlsanity.com/paper/2602.00722/) (2026)
- [Affordance-First Decomposition for Continual Learning in Video-Language Understanding](https://mlsanity.com/paper/2512.00694/) (2025)
- [LoRA-Loop: Closing the Synthetic Replay Cycle for Continual VLM Learning](https://mlsanity.com/paper/2507.13568/) (2025)
- [SITUATE -- Synthetic Object Counting Dataset for VLM training](https://mlsanity.com/paper/2602.00108/) (2026)
- [Dance Across Shifts: Forward-Facilitation Continual Test-Time Adaptation through Dynamic Style Bridging](https://mlsanity.com/paper/2605.18608/) (2026)
- [Physics-Based Benchmarking Metrics for Multimodal Synthetic Images](https://mlsanity.com/paper/2511.15204/) (2025)
- [Forging a Dynamic Memory: Retrieval-Guided Continual Learning for Generalist Medical Foundation Models](https://mlsanity.com/paper/2512.13072/) (2025)
- [Continual Learning with Vision-Language Models via Semantic-Geometry Preservation](https://mlsanity.com/paper/2603.12055/) (2026)
- [Crafting Your Evolving Dreams: Concept-Incremental Versatile Customization](https://mlsanity.com/paper/2606.04797/) (2026)
- [Hidden Forgetting in Continual Multimodal Learning: When Accuracy Survives but Grounding Fails](https://mlsanity.com/paper/2607.02020/) (2026)
- [Annotation-Free Class-Incremental Learning](https://mlsanity.com/paper/2511.19344/) (2025)
- [Synthetic Vasculature and Pathology Enhance Vision-Language Model Reasoning](https://mlsanity.com/paper/2512.11060/) (2025)
- [VisionFoundry: Teaching VLMs Visual Perception with Synthetic Images](https://mlsanity.com/paper/2604.09531/) (2026)
- [Enhancing Continual Learning of Vision-Language Models via Dynamic Prefix Weighting](https://mlsanity.com/paper/2604.18075/) (2026)
- [DIMoE-Adapters: Dynamic Expert Evolution for Continual Learning in Vision-Language Models](https://mlsanity.com/paper/2605.07494/) (2026)
- [iGSP:Implicit Gradient Subspace Projection for Efficient Continual Learning of Vision-Language Models](https://mlsanity.com/paper/2605.19301/) (2026)
- [KeepLoRA++: Continual Learning with Layer-Scaled Residual Gradient Adaptation](https://mlsanity.com/paper/2606.16256/) (2026)
- [InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories](https://mlsanity.com/paper/2607.02010/) (2026)

## Abstract (arXiv)

Pre-trained Vision-Language Models (VLMs) require Continual Learning (CL) to efficiently update their knowledge and adapt to various downstream tasks without retraining from scratch. However, for VLMs, in addition to the loss of knowledge previously learned from downstream tasks, pre-training knowledge is also corrupted during continual fine-tuning. This issue is exacerbated by the unavailability of original pre-training data, leaving VLM's generalization ability degrading. In this paper, we propose GIFT, a novel continual fine-tuning approach that utilizes synthetic data to overcome catastrophic forgetting in VLMs. Taking advantage of recent advances in text-to-image synthesis, we employ a pre-trained diffusion model to recreate both pre-training and learned downstream task data. In this way, the VLM can revisit previous knowledge through distillation on matching diffusion-generated images and corresponding text prompts. Leveraging the broad distribution and high alignment between synthetic image-text pairs in VLM's feature space, we propose a contrastive distillation loss along with an image-text alignment constraint. To further combat in-distribution overfitting and enhance distillation performance with limited amount of generated data, we incorporate adaptive weight consolidation, utilizing Fisher information from these synthetic image-text pairs and achieving a better stability-plasticity balance. Extensive experiments demonstrate that our method consistently outperforms previous state-of-the-art approaches across various settings.