Unverified50% confidenceFactExact time
The reference gap is a bottleneck independent of visual perception, explaining some errors previously misjudged as insufficient perception.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe researchers define the failure mode in which models' language is not correctly bound to visual entities, regions, or locations during visual reasoning as the "reference gap."78% similarUnverified3D可视化存在固有认知缺陷，包括遮挡问题、距离失真以及精确数值读取困难72% similarUnverifiedVisual reasoning involves implicit intermediate steps and lacks clear verification methods, making it long-standing difficult to model.72% similarUnverifiedCross-space alignment causes information loss, because the features learned by different encoders differ in semantics and scale, forcing alignment requires an additional projection network, and the projection may smooth away high-frequency texture details.71% similarUnverifiedThe essence of the reference gap is the manifestation of the coreference resolution problem in visual reasoning, where models are prone to referential drift during multi-step reasoning.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1113902API
curl https://kongchang.com/api/v1/knowledge/claims/1113902MCP
get_claim(id=1113902)