Unverified50% confidenceConceptExact time
The root of the modality accessibility gap can be traced to the quality of representation alignment: in UMMs, even if image generation and text question answering share backbone parameters, their downstream computation paths may still project knowledge into different subspaces.
1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn cross-modal retrieval, text vectors and image vectors still cluster in separate regions within the shared space, causing overall similarity scores between text and images to be systematically lower—a phenomenon known as the Modality Gap.75% similarUnverifiedThe study found a positive correlation between correct image reproduction and text accessibility, but a substantial gap still exists between the two (the modality accessibility gap).74% similarUnverifiedProcedurally generated data faces the domain gap problem: the textures, lighting, and object distributions of synthetic scenes differ from the real world, and models may learn surface statistical patterns rather than generalizable spatial reasoning ability.70% similarUnverified域间隙（Domain Gap）是合成数据领域的核心挑战，指渲染图像与真实相机图像在纹理细节、传感器噪声和光照物理特性等方面存在差距69% similarUnverifiedDomain Gap refers to the distribution difference between simulated images and real images.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1046218API
curl https://kongchang.com/api/v1/knowledge/claims/1046218MCP
get_claim(id=1046218)