# Multi-Label Contrastive Learning for Abstract Visual Reasoning

arXiv: https://arxiv.org/abs/2012.01944 · HTML: https://mlsanity.com/paper/2012.01944/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Dec 2020.

## Built on by (most cited first, 24 of 24)

- [Use All The Labels: A Hierarchical Multi-Label Contrastive Learning Framework](https://mlsanity.com/paper/2204.13207/) (2022)
- [Deep Learning Methods for Abstract Visual Reasoning: A Survey on Raven's Progressive Matrices](https://mlsanity.com/paper/2201.12382/) (2022)
- [A Review of Emerging Research Directions in Abstract Visual Reasoning](https://mlsanity.com/paper/2202.10284/) (2022)
- [Gaussian Mixture Variational Autoencoder with Contrastive Learning for Multi-Label Classification](https://mlsanity.com/paper/2112.00976/) (2021)
- [Evaluating Understanding on Conceptual Abstraction Benchmarks](https://mlsanity.com/paper/2206.14187/) (2022)
- [An Effective Deployment of Contrastive Learning in Multi-label Text Classification](https://mlsanity.com/paper/2212.00552/) (2022)
- [End-to-End Supervised Multilabel Contrastive Learning](https://mlsanity.com/paper/2307.03967/) (2023)
- [Bongard-RWR+: Real-World Representations of Fine-Grained Concepts in Bongard Problems](https://mlsanity.com/paper/2508.12026/) (2025)
- [Grounding and Distinguishing Conceptual Vocabulary Through Similarity Learning in Embodied Simulations](https://mlsanity.com/paper/2305.13668/) (2023)
- [One Self-Configurable Model to Solve Many Abstract Visual Reasoning Problems](https://mlsanity.com/paper/2312.09997/) (2023)
- [A-I-RAVEN and I-RAVEN-Mesh: Two New Benchmarks for Abstract Visual Reasoning](https://mlsanity.com/paper/2406.11061/) (2024)
- [Learning Visual Abstract Reasoning through Dual-Stream Networks](https://mlsanity.com/paper/2411.19451/) (2024)
- [Learning Label Hierarchy with Supervised Contrastive Learning](https://mlsanity.com/paper/2402.00232/) (2024)
- [Label Structure Preserving Contrastive Embedding for Multi-Label Learning with Missing Labels](https://mlsanity.com/paper/2209.01314/) (2022)
- [Computational Models of Solving Raven's Progressive Matrices: A Comprehensive Introduction](https://mlsanity.com/paper/2302.04238/) (2023)
- [Learning Abstract Visual Reasoning via Task Decomposition: A Case Study in Raven Progressive Matrices](https://mlsanity.com/paper/2308.06528/) (2023)
- [Systematic Abductive Reasoning via Diverse Relation Representations in Vector-symbolic Architecture](https://mlsanity.com/paper/2501.11896/) (2025)
- [Multi-level Supervised Contrastive Learning](https://mlsanity.com/paper/2502.02202/) (2025)
- [Coarse-to-Fine Contrastive Learning on Graphs](https://mlsanity.com/paper/2212.06423/) (2022)
- [A Feature-based Generalizable Prediction Model for Both Perceptual and Abstract Reasoning](https://mlsanity.com/paper/2403.05641/) (2024)
- [Computer User Interface Understanding. A New Dataset and a Learning Framework](https://mlsanity.com/paper/2403.10170/) (2024)
- [GenVP: Generating Visual Puzzles with Contrastive Hierarchical VAEs](https://mlsanity.com/paper/2503.23598/) (2025)
- [Advancing Generalization Across a Variety of Abstract Visual Reasoning Tasks](https://mlsanity.com/paper/2505.13391/) (2025)
- [Predictive Reasoning with Augmented Anomaly Contrastive Learning for Compositional Visual Relations](https://mlsanity.com/paper/2603.01125/) (2026)

## Abstract (arXiv)

For a long time the ability to solve abstract reasoning tasks was considered one of the hallmarks of human intelligence. Recent advances in application of deep learning (DL) methods led, as in many other domains, to surpassing human abstract reasoning performance, specifically in the most popular type of such problems - the Raven's Progressive Matrices (RPMs). While the efficacy of DL systems is indeed impressive, the way they approach the RPMs is very different from that of humans. State-of-the-art systems solving RPMs rely on massive pattern-based training and sometimes on exploiting biases in the dataset, whereas humans concentrate on identification of the rules / concepts underlying the RPM (or generally a visual reasoning task) to be solved. Motivated by this cognitive difference, this work aims at combining DL with human way of solving RPMs and getting the best of both worlds. Specifically, we cast the problem of solving RPMs into multi-label classification framework where each RPM is viewed as a multi-label data point, with labels determined by the set of abstract rules underlying the RPM. For efficient training of the system we introduce a generalisation of the Noise Contrastive Estimation algorithm to the case of multi-label samples. Furthermore, we propose a new sparse rule encoding scheme for RPMs which, besides the new training algorithm, is the key factor contributing to the state-of-the-art performance. The proposed approach is evaluated on two most popular benchmark datasets (Balanced-RAVEN and PGM) and on both of them demonstrates an advantage over the current state-of-the-art results. Contrary to applications of contrastive learning methods reported in other domains, the state-of-the-art performance reported in the paper is achieved with no need for large batch sizes or strong data augmentation.