Proceedings

Advancing Generalization Across a Variety of Abstract Visual Reasoning Tasks

Mikołaj Małkiński, Jacek Mańdziuk

Download BibTeX Cite

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. Main Track. Pages 5941-5949.

https://doi.org/10.24963/ijcai.2025/661

Keywords

  • Machine Learning: ML: Relational learning
  • Machine Learning: ML: Deep learning architectures
  • Machine Learning: ML: Multi-task and transfer learning

Abstract

The abstract visual reasoning (AVR) domain presents a diverse suite of analogy-based tasks devoted to studying model generalization. Recent years have brought dynamic progress in the field, particularly in i.i.d. scenarios, in which models are trained and evaluated on the same data distributions. Nevertheless, o.o.d. setups that assess model generalization to new test distributions remain challenging even for the most recent models. To advance generalization in AVR tasks, we present the Pathways of Normalized Group Convolution model (PoNG), a novel neural architecture that features group convolution, normalization, and a parallel design. We consider a wide set of AVR benchmarks, including Raven's Progressive Matrices and visual analogy problems with both synthetic and real-world images. The experiments demonstrate strong generalization capabilities of the proposed model, which in several settings outperforms the existing literature methods.

Mikołaj Małkiński, and Jacek Mańdziuk. 2025. Advancing Generalization Across a Variety of Abstract Visual Reasoning Tasks. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25. IJCAI, 5941–5949. https://doi.org/10.24963/ijcai.2025/661
Welcome Future Conferences AI Journal Division IJCAI Proceedings About IJCAI IJCAI Awards