# Directional Gradient Projection for Robust Fine-Tuning of Foundation Models

arXiv: https://arxiv.org/abs/2502.15895 · HTML: https://mlsanity.com/paper/2502.15895/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 29 of its 35 references that are in the map.
- First posted on arXiv 21 Feb 2025.

## Builds on (most cited first, 29 of 29)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [Towards VQA Models That Can Read](https://mlsanity.com/paper/1904.08920/) (2019)
- [An Empirical Study of Training Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.02057/) (2021)
- [Moment Matching for Multi-Source Domain Adaptation](https://mlsanity.com/paper/1812.01754/) (2018)
- [Gradient Surgery for Multi-Task Learning](https://mlsanity.com/paper/2001.06782/) (2020)
- [OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge](https://mlsanity.com/paper/1906.00067/) (2019)
- [EVA-CLIP: Improved Training Techniques for CLIP at Scale](https://mlsanity.com/paper/2303.15389/) (2023)
- [Robust fine-tuning of zero-shot models](https://mlsanity.com/paper/2109.01903/) (2021)
- [Fine-Tuning can Distort Pretrained Features and Underperform Out-of-Distribution](https://mlsanity.com/paper/2202.10054/) (2022)
- [Don't Just Assume; Look and Answer: Overcoming Priors for Visual Question Answering](https://mlsanity.com/paper/1712.00377/) (2017)
- [Explicit Inductive Bias for Transfer Learning with Convolutional Networks](https://mlsanity.com/paper/1802.01483/) (2018)
- [Finetune like you pretrain: Improved finetuning of zero-shot vision models](https://mlsanity.com/paper/2212.00638/) (2022)
- [Towards Causal VQA: Revealing and Reducing Spurious Correlations by Invariant and Covariant Semantic Editing](https://mlsanity.com/paper/1912.07538/) (2019)
- [Cycle-Consistency for Robust Visual Question Answering](https://mlsanity.com/paper/1902.05660/) (2019)
- [Beyond Question-Based Biases: Assessing Multimodal Shortcut Learning in Visual Question Answering](https://mlsanity.com/paper/2104.03149/) (2021)
- [Human-Adversarial Visual Question Answering](https://mlsanity.com/paper/2106.02280/) (2021)
- [Distance-Based Regularisation of Deep Networks for Fine-Tuning](https://mlsanity.com/paper/2002.08253/) (2020)
- [A Closer Look at the Robustness of Vision-and-Language Pre-trained Models](https://mlsanity.com/paper/2012.08673/) (2020)
- [Trainable Projected Gradient Method for Robust Fine-tuning](https://mlsanity.com/paper/2303.10720/) (2023)
- [Open-ended VQA benchmarking of Vision-Language models by exploiting Classification datasets and their semantic hierarchy](https://mlsanity.com/paper/2402.07270/) (2024)
- [C-VQA: A Compositional Split of the Visual Question Answering (VQA) v1.0 Dataset](https://mlsanity.com/paper/1704.08243/) (2017)
- [Robust Visual Question Answering: Datasets, Methods, and Future Challenges](https://mlsanity.com/paper/2307.11471/) (2023)
- [Reassessing Evaluation Practices in Visual Question Answering: A Case Study on Out-of-Distribution Generalization](https://mlsanity.com/paper/2205.12191/) (2022)
- [Fast Trainable Projection for Robust Fine-Tuning](https://mlsanity.com/paper/2310.19182/) (2023)
- [SAFT: Towards Out-of-Distribution Generalization in Fine-Tuning](https://mlsanity.com/paper/2407.03036/) (2024)
- [Rethinking Weight Decay for Robust Fine-Tuning of Foundation Models](https://mlsanity.com/paper/2411.01713/) (2024)

## Built on by (most cited first, 11 of 11)

- [Learning to Unlearn while Retaining: Combating Gradient Conflicts in Machine Unlearning](https://mlsanity.com/paper/2503.06339/) (2025)
- [Gradient Surgery for Safe LLM Fine-Tuning](https://mlsanity.com/paper/2508.07172/) (2025)
- [Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint](https://mlsanity.com/paper/2509.06795/) (2025)
- [MAPS: Preserving Vision-Language Representations via Module-Wise Proximity Scheduling for Better Vision-Language-Action Generalization](https://mlsanity.com/paper/2511.19878/) (2025)
- [FRAMES-VQA: Benchmarking Fine-Tuning Robustness across Multi-Modal Shifts in Visual Question Answering](https://mlsanity.com/paper/2505.21755/) (2025)
- [AEGIS: Adversarial Target-Guided Retention-Data-Free Robust Concept Erasure from Diffusion Models](https://mlsanity.com/paper/2602.06771/) (2026)
- [Deep Learning for Crack Detection: A Review of Learning Paradigms, Generalizability, and Datasets](https://mlsanity.com/paper/2508.10256/) (2025)
- [LARGO: Low-Rank Regulated Gradient Projection for Robust Parameter Efficient Fine-Tuning](https://mlsanity.com/paper/2506.12394/) (2025)
- [The Geometry of Robustness: Optimizing Loss Landscape Curvature and Feature Manifold Alignment for Robust Finetuning of Vision-Language Models](https://mlsanity.com/paper/2603.27139/) (2026)
- [The Devil Is in Gradient Entanglement: Energy-Aware Gradient Coordinator for Robust Generalized Category Discovery](https://mlsanity.com/paper/2604.14176/) (2026)
- [Text-Centric Post-Training for Omni-Modal Reasoning](https://mlsanity.com/paper/2610.02819/) (2026)

## Abstract (arXiv)

Robust fine-tuning aims to adapt large foundation models to downstream tasks while preserving their robustness to distribution shifts. Existing methods primarily focus on constraining and projecting current model towards the pre-trained initialization based on the magnitudes between fine-tuned and pre-trained weights, which often require extensive hyper-parameter tuning and can sometimes result in underfitting. In this work, we propose Directional Gradient Projection (DiGraP), a novel layer-wise trainable method that incorporates directional information from gradients to bridge regularization and multi-objective optimization. Besides demonstrating our method on image classification, as another contribution we generalize this area to the multi-modal evaluation settings for robust fine-tuning. Specifically, we first bridge the uni-modal and multi-modal gap by performing analysis on Image Classification reformulated Visual Question Answering (VQA) benchmarks and further categorize ten out-of-distribution (OOD) VQA datasets by distribution shift types and degree (i.e. near versus far OOD). Experimental results show that DiGraP consistently outperforms existing baselines across Image Classfication and VQA tasks with discriminative and generative backbones, improving both in-distribution (ID) generalization and OOD robustness.