analiza aiqbench
The paper introduces ImagiRep, a benchmark that tests whether vision-language models reason better about unseen scenes when given depth, panoramic, or top-down representations instead of RGB images. Its main finding is that accurate imagined RGB doesn't guarantee correct spatial answers, and that the best representation depends on the question. For researchers building world models, this suggests the generated output format may be a bottleneck worth optimizing separately from image quality.
Representing the Unseen: Rethinking Visual Imagination for Spatial Reasoning
arXiv:2610.37437v1 Announce Type: cross Abstract: Visual world models increasingly support spatial reasoning by generating RGB views of unobserved states. However, RGB has become the default interface largely because it is well suited for visual generation, rather than necessarily being the most effective representation for downstream reasoning. Different spatial questions may benefit from different representations of the same imagined state: top-down geometry can facilitate global distance comparisons, while panoramic views provide broader context for reasoning about open space. To isolate the role of representation in spatial imagination, we introduce ImagiRep, a simulator-grounded benchmark of 922 questions spanning seven question types across two settings, Viewpoint Imagination and Cross-view Integration, providing oracle RGB and six alternative representations (e.g., depth, panoramic views, and top-down geometry) of the same unobserved state. This controlled comparison reveals that even accurate imagined RGB does not guarantee accurate spatial reasoning, and different questions benefit from different representations. Similar trends hold for representations derived from world-model-estimated states. Notably, estimated RGB alone can mislead spatial reasoning, while alternative representations retain complementary benefits. Building on these findings, we introduce GEARS, a lightweight test-time strategy for adaptive representation selection, outperforming existing spatial imagination methods and showing that reasoning about unseen states requires more than generating RGB views.
Abstract:Visual world models increasingly support spatial reasoning by generating RGB views of unobserved states. However, RGB has become the default interface largely because it is well suited for visual generation, rather than necessarily being the most effective representation for downstream reasoning. Different spatial questions may benefit from different representations of the same imagined state: top-down geometry can facilitate global distance comparisons, while panoramic views provide broader context for reasoning about open space. To isolate the role of representation in spatial imagination, we introduce ImagiRep, a simulator-grounded benchmark of 922 questions spanning seven question types across two settings, Viewpoint Imagination and Cross-view Integration, providing oracle RGB and six alternative representations (e.g., depth, panoramic views, and top-down geometry) of the same unobserved state. This controlled comparison reveals that even accurate imagined RGB does not guarantee accurate spatial reasoning, and different questions benefit from different representations. Similar trends hold for representations derived from world-model-estimated states. Notably, estimated RGB alone can mislead spatial reasoning, while alternative representations retain complementary benefits. Building on these findings, we introduce GEARS, a lightweight test-time strategy for adaptive representation selection, outperforming existing spatial imagination methods and showing that reasoning about unseen states requires more than generating RGB views.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2610.37437 [cs.CV]
(or arXiv:2610.37437v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2610.37437 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 303 cuvinte
Subiecte conexe
ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning
arXiv · 29 oct. 2026
Contemporary AI lacks the imagination to diverge or negate in science
arXiv · 10 iun. 2026
Contemporary AI lacks the imagination to diverge or negate in science
arXiv · 9 iun. 2026
Researchers Link Dimensional Codes to Toric Code Families
Quantum Zeitgeist · 24 aug. 2026