Abstract Visual Reasoning with Tangram Shapes
This work addresses the challenge of evaluating and improving abstract visual reasoning in AI systems, which is incremental as it builds on existing cognitive science stimuli and multi-modal models.
The authors tackled the problem of abstract visual reasoning by introducing KiloGram, a large and diverse dataset of tangram shapes, and found that fine-tuning multi-modal models on it dramatically improved their reasoning abilities, with explicit part descriptions further enhancing performance for both humans and models.
We introduce KiloGram, a resource for studying abstract visual reasoning in humans and machines. Drawing on the history of tangram puzzles as stimuli in cognitive science, we build a richly annotated dataset that, with >1k distinct stimuli, is orders of magnitude larger and more diverse than prior resources. It is both visually and linguistically richer, moving beyond whole shape descriptions to include segmentation maps and part labels. We use this resource to evaluate the abstract visual reasoning capacities of recent multi-modal models. We observe that pre-trained weights demonstrate limited abstract reasoning, which dramatically improves with fine-tuning. We also observe that explicitly describing parts aids abstract reasoning for both humans and models, especially when jointly encoding the linguistic and visual inputs. KiloGram is available at https://lil.nlp.cornell.edu/kilogram .