CogUtil::SIMDOptimizations
SIMD-optimized vector operations for cognitive tensors
Class methods
apply_attention_weights(tensor : Array(Float32), weights : Array(Float32)) : Array(Float32)
Vectorized attention weight application
dot_product(a : Array(Float32), b : Array(Float32)) : Float32
Vectorized dot product for cognitive attention calculations
normalize_l2(vector : Array(Float32)) : Array(Float32)
Fast normalization for cognitive tensors