Aug 2026· Science Advances· Vol 12· 0 citations· 57 references
Medicine
TL;DR
A self-supervised framework that trains models on systematically simplified versions of abstract reasoning tasks containing incomplete but structured concept cues enables models to form internal abstractions under limited resources and later apply them to more complex problems.
Abstract
How agents acquire abstract concepts from sparse, diverse examples—often without explicit supervision—remains a central problem in cognitive science and artificial intelligence. Human studies suggest that this ability depends on mental bootstrapping, the gradual construction of complex concepts from simpler partial structures. Building on this idea, we develop a self-supervised framework that trains models on systematically simplified versions of abstract reasoning tasks containing incomplete but structured concept cues. This algorithm enables models to form internal abstractions under limited resources and later apply them to more complex problems. We evaluate the framework across 12 abstract visual reasoning datasets testing in-distribution concept induction, out-of-distribution generalization, and few-shot learning. To contextualize performance, we also measure human accuracy on the same tasks. Models trained on simplified problems generalize robustly, reaching or even surpassing human-level performance. These findings show that abstract reasoning can emerge from structured simplification and minimal data, offering a computational account of concept learning in humans and machines.
A dependency-graph framework is proposed to formalize compositional reasoning, yielding three levels of compositionality with increasing complexity, and preliminary evidence that the decomposed-to-composed asymmetry can extend to practical settings is presented.
Yu He, Ying-Xi Li, Yi-Fei Wang et al.· 0 citations
A scalable hierarchical multimodal recurrent neural network grounded in predictive processing under the free-energy principle, capable of directly integrating more than 30,000-dimensional visuo-proprioceptive inputs without dimensionality reduction or handcrafted preprocessing is introduced.
This analysis provides a structured account of current approaches to scaling LRMs beyond human supervision and the open problems involved in developing self-sustaining learning systems toward superintelligence.
Zhiqin Yang, Jing-Wen Fu, Yu-Han Liu et al.· 4 citations
Overall, Neural -Symbolic AI shifts computer vision from pure perception toward structured understanding and reasoning, although problems remain in scalable reasoning, knowledge integration, and real-world deployment.
Func-R1 is proposed, which synergistically harmonizes precise visual perception and rigorous logical reasoning and delivers the optimal performance among open-source MLLMs, even surpassing GPT-5 with an 8.4% improvement on MathVerse's function-oriented tasks.
Ming Yin, Xiaohai Wang, Dian Li et al.· Conference on Empirical Meth...· 1 citation
Visual Question Answering (VQA) involves models combining reasoning through visual scenes and natural language questions and typically related to compositional and relational reasoning. Even though deep neural models have demonstrated high empirical results on VQA benchmarks, they are often based on implicit associatio...
Akash Badhan, Priyank Arora, Rishabh Garg et al.· International Conference on...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.