# An Attention Mechanism for Robust Multimodal Integration in a Global Workspace Architecture

arXiv: https://arxiv.org/abs/2602.08597 · HTML: https://mlsanity.com/paper/2602.08597/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 31 references that are in the map.
- First posted on arXiv 9 Feb 2026.

## Builds on (most cited first, 23 of 23)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021)
- [Energy and Policy Considerations for Deep Learning in NLP](https://mlsanity.com/paper/1906.02243/) (2019)
- [Multimodal Machine Learning: A Survey and Taxonomy](https://mlsanity.com/paper/1705.09406/) (2017)
- [Multimodal Transformer for Unaligned Multimodal Language Sequences](https://mlsanity.com/paper/1906.00295/) (2019)
- [Language Is Not All You Need: Aligning Perception with Language Models](https://mlsanity.com/paper/2302.14045/) (2023)
- [Multimodal Deep Learning](https://mlsanity.com/paper/2301.04856/) (2023)
- [Gated Multimodal Units for Information Fusion](https://mlsanity.com/paper/1702.01992/) (2017)
- [SMIL: Multimodal Learning with Severely Missing Modality](https://mlsanity.com/paper/2103.05677/) (2021)
- [Unsupervised Neural Machine Translation](https://mlsanity.com/paper/1710.11041/) (2017)
- [Phrase-Based & Neural Unsupervised Machine Translation](https://mlsanity.com/paper/1804.07755/) (2018)
- [Deep Multimodal Learning with Missing Modality: A Survey](https://mlsanity.com/paper/2409.07825/) (2024)
- [Coordination Among Neural Modules Through a Shared Global Workspace](https://mlsanity.com/paper/2103.01197/) (2021)
- [MFAS: Multimodal Fusion Architecture Search](https://mlsanity.com/paper/1903.06496/) (2019)
- [What to align in multimodal contrastive learning?](https://mlsanity.com/paper/2409.07402/) (2024)
- [Deep Learning and the Global Workspace Theory](https://mlsanity.com/paper/2012.10390/) (2020)
- [Benchmarking Robustness of Multimodal Image-Text Models under Distribution Shift](https://mlsanity.com/paper/2212.08044/) (2022)
- [Semi-supervised Multimodal Representation Learning through a Global Workspace](https://mlsanity.com/paper/2306.15711/) (2023)
- [Associative Transformer](https://mlsanity.com/paper/2309.12862/) (2023)
- [Incorporating Domain Knowledge Graph into Multimodal Movie Genre Classification with Self-Supervised Attention and Contrastive Learning](https://mlsanity.com/paper/2310.08032/) (2023)
- [Zero-shot cross-modal transfer of Reinforcement Learning policies through a Global Workspace](https://mlsanity.com/paper/2403.04588/) (2024)
- [Multimodal Dreaming: A Global Workspace Approach to World Model-Based Reinforcement Learning](https://mlsanity.com/paper/2502.21142/) (2025)

## Abstract (arXiv)

Robust multimodal systems must remain effective when some modalities are noisy, degraded, or unreliable. Existing multimodal fusion methods often learn modality selection jointly with representation learning, making it difficult to determine whether robustness comes from the selector itself or from full end-to-end co-adaptation. Motivated by Global Workspace Theory (GWT), we study this question using a lightweight top-down modality selector operating on top of a frozen multimodal global workspace. We evaluate our method on two multimodal datasets of increasing complexity: Simple Shapes and MM-IMDb 1.0, under structured modality corruptions. The selector improves robustness while using far fewer trainable parameters than end-to-end attention baselines, and the learned selection strategy transfers better across downstream tasks, corruption regimes, and even to a previously unseen modality. Beyond explicit corruption settings, on the MM-IMDb 1.0 benchmark, we show that the same mechanism improves the global workspace over its no-attention counterpart and yields decent benchmark performance.