Unverified50% confidenceFactExact time
In SpecSlot, only the lowest-frequency slot of each anchor participates in global attention, while all slots interact locally, so increasing spectral detail (the number of slots) does not cause the global attention sequence length to expand.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedExisting mainstream sparse attention methods filter interactions by retaining entries with larger values or high-quality regions in the attention matrix.72% similarUnverifiedIQuest-Q1 employs a mixed ratio of 3 layers of sliding window attention combined with 1 layer of full attention, with only 25 out of 88 layers expanding the KV cache as context grows.71% similarUnverifiedDense attention layers require a complete KV Cache, whereas linear layers rely on hidden states.71% similarUnverifiedAs the number of tokens in the window grows, attention weights must be distributed among these tokens, and the average attention each individual token receives declines accordingly.70% similarUnverified自注意力机制的计算路径长度与序列长度无关，任意两个位置之间的信息传递都是直接的，解决了长序列建模中的梯度消失问题70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1115713API
curl https://kongchang.com/api/v1/knowledge/claims/1115713MCP
get_claim(id=1115713)