Expired50% confidenceOpinionExact time
Although large text models have surpassed humans on various text tasks, multimodal large models still significantly lag behind human performance on visual tasks.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/16/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/14/2026(expired)
Sources
Related Claims
UnverifiedMultimodal large models can leverage language priors to complete ambiguous characters, greatly diminishing the effectiveness of attacks targeting the visual features of individual characters.67% similarUnverified对于需要精确视觉推理的场景（文档理解、图表问答、机器人视觉、医学影像），单靠CLIP视觉编码器不够，建议在视觉编码器层面做增强而非仅依赖更大的语言模型66% similarUnverified现实世界的细节远比人们头脑中的抽象模型所能容纳的要多得多66% similarUnverifiedDOM裁剪速度快但可能丢失视觉布局信息，可访问性树语义准确但对动态渲染内容覆盖不完整，VLM方案最接近人类视觉理解但推理延迟高、成本大64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/533781API
curl https://kongchang.com/api/v1/knowledge/claims/533781MCP
get_claim(id=533781)