Module binary 

Binary Quantization — 32x compression (1 bit per dimension).

Used for the Cold (Tier 2) tier. Encodes only the sign of each dimension and uses Hamming distance for comparison.

Functions

decode_binary

Decode binary codes back to an approximate float vector.

encode_binary

Encode a float vector to binary: 1 bit per dimension (sign bit).

hamming_distance

Compute the Hamming distance between two binary-encoded vectors.