Unverified50% confidenceFactExact time
DeltaNet is a variant of linear attention that replaces softmax attention with a delta update rule, reducing sequence processing complexity from O(n²) to O(n).
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe time complexity of the Leiden algorithm is approximately linear (O(n log n)), which gives it good scalability in engineering applications.58% similarUnverifiedRecurrentGemma adopts a Linear Recurrent Unit (LRU) architecture, reducing the computational complexity of attention from quadratic growth with sequence length to linear complexity.57% similarUnverifiedSpeculative decoding improves inference speed by using a small draft model to generate candidate tokens that are verified in parallel by the main model; sparse attention reduces computational complexity from O(n²) to near O(n).55% similarUnverifiedThe core processes of LLMs include tokenization, embedding layer, self-attention mechanism, feed-forward network, and residual connections.55% similarUnverifiedThe prefill stage is compute-bound, requiring all input tokens to be processed at once to generate the KV Cache, with attention matrix computation complexity of O(n²).54% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/896121API
curl https://kongchang.com/api/v1/knowledge/claims/896121MCP
get_claim(id=896121)