Unverified50% confidenceFactExact time
标准Transformer的单次前向传播等价于固定深度的计算图，其表达能力受限于TC^0复杂度类，但通过多步推理可接近图灵完备的计算能力
1
Sources
50%
Confidence
Long-term
Relevance
8/20/2026
First Seen
Sources
Related Claims
VerifiedTransformer overcomes the long-range dependency attenuation caused by sequential computation in RNN/LSTM architectures, while natively supporting parallel computation.73% similarUnverifiedSet Transformer reduces complexity to O(mn) by introducing learnable Inducing Points.73% similarUnverifiedTransformer 中任意两个位置之间的信息传递路径长度恒为1，从根本上克服了 RNN 在长序列上因梯度消失导致的遗忘问题71% similarUnverifiedSSMs have a linear time complexity advantage in processing long sequences, whereas the attention computation complexity of standard Transformers is quadratic with respect to sequence length.69% similarVerifiedTransformer架构的自注意力机制在处理长序列时面临二次方复杂度的计算开销，上下文长度翻倍则计算量增长四倍69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/779383API
curl https://kongchang.com/api/v1/knowledge/claims/779383MCP
get_claim(id=779383)