# Mirror Gradient: Towards Robust Multimodal Recommender Systems via Exploring Flat Local Minima

arXiv: https://arxiv.org/abs/2402.11262 · HTML: https://mlsanity.com/paper/2402.11262/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 36 of its 65 references that are in the map.
- First posted on arXiv 17 Feb 2024.

## Builds on (most cited first, 30 of 36)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Sharpness-Aware Minimization for Efficiently Improving Generalization](https://mlsanity.com/paper/2010.01412/) (2020)
- [LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation](https://mlsanity.com/paper/2002.02126/) (2020)
- [Visualizing the Loss Landscape of Neural Nets](https://mlsanity.com/paper/1712.09913/) (2017)
- [Image-based Recommendations on Styles and Substitutes](https://mlsanity.com/paper/1506.04757/) (2015)
- [Generating Sequences With Recurrent Neural Networks](https://mlsanity.com/paper/1308.0850/) (2013)
- [ASAM: Adaptive Sharpness-Aware Minimization for Scale-Invariant Learning of Deep Neural Networks](https://mlsanity.com/paper/2102.11600/) (2021)
- [Bias and Debias in Recommender System: A Survey and Future Directions](https://mlsanity.com/paper/2010.03240/) (2020)
- [VBPR: Visual Bayesian Personalized Ranking from Implicit Feedback](https://mlsanity.com/paper/1510.01784/) (2015)
- [Efficient Sharpness-aware Minimization for Improved Training of Neural Networks](https://mlsanity.com/paper/2110.03141/) (2021)
- [Penalizing Gradient Norm for Efficiently Improving Generalization in Deep Learning](https://mlsanity.com/paper/2202.03599/) (2022)
- [Overcoming Catastrophic Forgetting in Incremental Few-Shot Learning by Finding Flat Minima](https://mlsanity.com/paper/2111.01549/) (2021)
- [Collaborative Deep Learning for Recommender Systems](https://mlsanity.com/paper/1409.2944/) (2014)
- [Asymmetric Valleys: Beyond Sharp and Flat Local Minima](https://mlsanity.com/paper/1902.00744/) (2019)
- [When Do Flat Minima Optimizers Work?](https://mlsanity.com/paper/2202.00661/) (2022)
- [Make Sharpness-Aware Minimization Stronger: A Sparsified Perturbation Approach](https://mlsanity.com/paper/2210.05177/) (2022)
- [Ranking Distillation: Learning Compact Ranking Models With High Performance for Recommender System](https://mlsanity.com/paper/1809.07428/) (2018)
- [GRCN: Graph-Refined Convolutional Network for Multimedia Recommendation with Implicit Feedback](https://mlsanity.com/paper/2111.02036/) (2021)
- [Let's Think Outside the Box: Exploring Leap-of-Thought in Large Language Models with Creative Humor Generation](https://mlsanity.com/paper/2312.02439/) (2023)
- [Bootstrap Latent Representations for Multi-modal Recommendation](https://mlsanity.com/paper/2207.05969/) (2022)
- [A Tale of Two Graphs: Freezing and Denoising Graph Structures for Multimodal Recommendation](https://mlsanity.com/paper/2211.06924/) (2022)
- [DIANet: Dense-and-Implicit Attention Network](https://mlsanity.com/paper/1905.10671/) (2019)
- [Convolution-Weight-Distribution Assumption: Rethinking the Criteria of Channel Pruning](https://mlsanity.com/paper/2004.11627/) (2020)
- [SUR-adapter: Enhancing Text-to-Image Pre-trained Diffusion Models with Large Language Models](https://mlsanity.com/paper/2305.05189/) (2023)
- [Adversarial Training Towards Robust Multimedia Recommender System](https://mlsanity.com/paper/1809.07062/) (2018)
- [Instance Enhancement Batch Normalization: an Adaptive Regulator of Batch Noise](https://mlsanity.com/paper/1908.04008/) (2019)
- [Dynamically Expandable Graph Convolution for Streaming Recommendation](https://mlsanity.com/paper/2303.11700/) (2023)
- [ScaleLong: Towards More Stable Training of Diffusion Model via Scaling Network Long Skip Connection](https://mlsanity.com/paper/2310.13545/) (2023)
- [SelfCF: A Simple Framework for Self-supervised Collaborative Filtering](https://mlsanity.com/paper/2107.03019/) (2021)
- [Enhancing Dyadic Relations with Homogeneous Graphs for Multimodal Recommendation](https://mlsanity.com/paper/2301.12097/) (2023)

## Built on by (most cited first, 6 of 6)

- [RouterEval: A Comprehensive Benchmark for Routing LLMs to Explore Model-level Scaling Up in LLMs](https://mlsanity.com/paper/2503.10657/) (2025)
- [A Causality-aware Paradigm for Evaluating Creativity of Multimodal Large Language Models](https://mlsanity.com/paper/2501.15147/) (2025)
- [Dreaming User Multimodal Representation Guided by The Platonic Representation Hypothesis for Micro-Video Recommendation](https://mlsanity.com/paper/2410.03538/) (2024)
- [Multimodal Representation-disentangled Information Bottleneck for Multimodal Recommendation](https://mlsanity.com/paper/2509.20225/) (2025)
- [Structurally Refined Graph Transformer for Multimodal Recommendation](https://mlsanity.com/paper/2511.00584/) (2025)
- [Anchored Alignment: Preventing Positional Collapse in Multimodal Recommender Systems](https://mlsanity.com/paper/2603.12726/) (2026)

## Abstract (arXiv)

Multimodal recommender systems utilize various types of information to model user preferences and item features, helping users discover items aligned with their interests. The integration of multimodal information mitigates the inherent challenges in recommender systems, e.g., the data sparsity problem and cold-start issues. However, it simultaneously magnifies certain risks from multimodal information inputs, such as information adjustment risk and inherent noise risk. These risks pose crucial challenges to the robustness of recommendation models. In this paper, we analyze multimodal recommender systems from the novel perspective of flat local minima and propose a concise yet effective gradient strategy called Mirror Gradient (MG). This strategy can implicitly enhance the model's robustness during the optimization process, mitigating instability risks arising from multimodal information inputs. We also provide strong theoretical evidence and conduct extensive empirical experiments to show the superiority of MG across various multimodal recommendation models and benchmarks. Furthermore, we find that the proposed MG can complement existing robust training methods and be easily extended to diverse advanced recommendation models, making it a promising new and fundamental paradigm for training multimodal recommender systems. The code is released at https://github.com/Qrange-group/Mirror-Gradient.