Unverified50% confidenceOpinionExact time
In spiking Transformers, membrane variables are usually as important as weights, because the query, key, and value computations in the attention mechanism all depend on spike sequences, and small deviations in membrane variables propagate and accumulate across time steps, causing firing patterns to deviate.
1
Sources
50%
Confidence
Long-term
Relevance
11/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe Transformer's self-attention mechanism is more sensitive to memory bandwidth and capacity, and performance bottlenecks readily appear on hardware with limited memory.71% similarUnverifiedTransformer invariants include shape invariants, numerical properties (such as the sum of each row of softmax output always equaling 1 and attention weights being non-negative), and equivariance/invariance.65% similarUnverifiedLayers inside a Transformer vary greatly in their sensitivity to precision loss, and uniform quantization often pays too much quality cost on a few sensitive layers.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1182845API
curl https://kongchang.com/api/v1/knowledge/claims/1182845MCP
get_claim(id=1182845)