Expired50% confidenceFactExact time
SE (Squeeze-and-Excitation) and CBAM (Convolutional Block Attention Module) are common attention modules in the field of computer vision.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026(expired)
Sources
Related Claims
UnverifiedCBAM (Convolutional Block Attention Module) further introduces a spatial attention branch on top of SE-Net, attending to both channel importance and spatial location importance simultaneously.81% similarUnverifiedUltra-large context windows are typically implemented relying on sparse attention mechanisms, RoPE positional encoding extrapolation, and hierarchical caching technology.68% similarUnverified上下文窗口扩展得益于ALiBi位置编码、RoPE的外推扩展、Flash Attention和Ring Attention等技术突破64% similarUnverifiedLinformer 将注意力矩阵低秩近似为 O(n) 复杂度，FlashAttention 通过重排计算顺序降低 GPU 内存带宽占用，Longformer 引入滑动窗口注意力63% similarUnverified新内核为滑动窗口注意力（SWA）层提供双输入支持，并引入 attention sinks 机制62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/49252API
curl https://kongchang.com/api/v1/knowledge/claims/49252MCP
get_claim(id=49252)