mlsanityData to 14 Oct 2026

GraphSelect for Budgeted Representation Selection in Multimodal Graph Inference

arXiv 2609.33561 ↗First posted 27 Sep 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Multimodal graph predictors combine text, images, and relations to classify connected entities. How much of this input is needed to preserve their predictions? We study budgeted representation selection, which chooses a subset of candidate text and image vectors under a separate capacity for each modality. Predictions from the complete candidate input define the classes to preserve. The challenge is that a representation's contribution depends on the other selected inputs, while graph propagation extends its effects across nodes. Our empirical study shows that candidate rankings change with the selected input, while predicted probabilities remain informative after the class stops changing. Updating scores improves selection, and exchanging inputs can improve a subset whose capacity is already filled. These findings lead to GraphSelect, which starts from individual candidate gains and refines the subset through jointly evaluated exchanges. It screens promising removals and additions, accepts an exchange when it reduces the prediction loss, and updates the scores. Experiments on six graphs show higher mean objective recovery than six attribution and explanation methods adapted to the selection task. Across nine trained architectures on two graphs, retaining 20% of the candidate representations per modality gives a mean accuracy drop of 0.10 percentage points relative to full candidate input, preserving classification performance with substantially fewer text and image representations.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 31 of its 37 references that are in the map. First posted on arXiv 27 Sep 2026.

Builds on

31 of 37 references.

References · 31

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  2. 2017Graph Attention Networkscited by 8,923
  3. 2017Inductive Representation Learning on Large Graphscited by 6,320
  4. 2017Axiomatic Attribution for Deep Networkscited by 3,384
  5. 2019GNNExplainer: Generating Explanations for Graph Neural Networkscited by 737
  6. 2020Captum: A unified and generic model interpretability library for PyTorchcited by 547
  7. 2021On Explainability of Graph Neural Networks via Subgraph Explorationscited by 281
  8. 2021GRAD-MATCH: Gradient Matching based Data Subset Selection for Efficient Deep Model Trainingcited by 274
  9. 2020Graph Structured Network for Image-Text Matchingcited by 93
  10. 2022GraphFramEx: Towards Systematic Evaluation of Explainability Methods for Graph Neural Networkscited by 58
  11. 2022Multimodal learning with graphscited by 54
  12. 2025UniGraph2: Learning a Unified Embedding Space to Bind Multimodal Graphscited by 49
19 more
  1. 2023LGMRec: Local and Global Graph Learning for Multimodal Recommendationcited by 38
  2. 2021Zorro: Valid, Sparse, and Stable Explanations in Graph Neural Networkscited by 35
  3. 2018What made you do this? Understanding black-box decisions with sufficient input subsetscited by 31
  4. 2024Rethinking Data Shapley for Data Selection Tasks: Misleads and Meritscited by 29
  5. 2024Mosaic of Modalities: A Comprehensive Benchmark for Multimodal Graph Learningcited by 29
  6. 2023Multimodal Graph Learning for Generative Taskscited by 27
  7. 2024Learning on Multimodal Graphs: A Surveycited by 22
  8. 2024When Graph meets Multimodal: Benchmarking and Meditating on Multimodal Attributed Graphs Learningcited by 22
  9. 2025GRAPHGPT-O: Synergistic Multimodal Comprehension and Generation on Graphscited by 17
  10. 2025Graph4MM: Weaving Multimodal Learning with Structural Informationcited by 17
  11. 2025Cross-Contrastive Clustering for Multimodal Attributed Graphs with Dual Graph Filteringcited by 14
  12. 2026OpenMAG: A Comprehensive Benchmark for Multimodal-Attributed Graphcited by 13
  13. 2024GNNShap: Scalable and Accurate GNN Explanation using Shapley Valuescited by 12
  14. 2024GOAt: Explaining Graph Neural Networks via Graph Output Attributioncited by 10
  15. 2025Graph-MLLM: Harnessing Multimodal Large Language Models for Multimodal Graph Learningcited by 9
  16. 2025Cohort-Based Active Modality Acquisitioncited by 5
  17. 2022Greedy Modality Selection via Approximate Submodular Maximizationcited by 3
  18. 2026LION: A Clifford Neural Paradigm for Multimodal-Attributed Graph Learningcited by 2
  19. 2026Context-aware Modality-Topology Co-Alignment for Multimodal Attributed Graphscited by 1

Built on by

Not cited yet.