In-memory resonator network for factorizing hyper vectors

A method for factorizing an input hypervector comprising:

• storing respective code hypervectors of an associated codebook within row lines of a neuromorphic memory device with a crossbar array structure comprising the row lines and column lines and resistive memory elements coupled between the row lines and the column lines at junctions formed by the row and column lines;
• each associated codebook comprising candidate code hypervectors;
• storing an input hypervector in an input buffer;
• storing estimated hypervectors within a plurality of estimate buffers each associated with a different subset of row lines and a different codebook, wherein the estimated hypervectors are derived from respective code hypervectors that are stored in the associated subset of row lines;
• computing an unbound hypervector using the input hypervector and all the estimated hypervectors;
• computing an attention vector that indicates a similarity of the unbound hypervector with one estimated hypervector;
• computing a linear combination of the one estimated hypervector, weighted by the attention vector;
• and
• storing the computed linear combination in the estimate buffer that is associated with the one estimated hypervector.

A computer program product comprising a computer-readable storage medium having computer-readable program code embodied therewith, wherein when the computer-readable program code is evoked by a processor causes the processor to:

• store respective code hypervectors of an associated codebook within row lines of a neuromorphic memory device with a crossbar array structure comprising the row lines and column lines and resistive memory elements coupled between the row lines and the column lines at junctions formed by the row and column lines;
• each associated codebook comprising candidate code hypervectors;
• store an input hypervector in an input buffer;
• store estimated hypervectors within a plurality of estimate buffers each associated with a different subset of row lines and a different codebook, wherein the estimated hypervectors are derived from respective code hypervectors that are stored in the associated subset of row lines;
• compute an unbound hypervector using the input hypervector and all the estimated hypervectors;
• compute an attention vector that indicates a similarity of the unbound hypervector with one estimated hypervector;
• compute a linear combination of the one estimated hypervector, weighted by the attention vector;
• and
• store the computed linear combination in the estimate buffer that is associated with the one estimated hypervector.