QELM

Quantum-Enhanced Language Model, a public hybrid framework with trainable quantum channels, amplitude-and-phase features, attention and feed-forward paths, next-token training, saved-model tools, and QSA-backed exact execution.

Scope

Working areas

01Trainable quantum channels and sub-bit features
02Parameter-shift, SPSA, and Adam training
03QSA, CPU, Aer, and IBM-compatible routes
04Trainer and saved-model chat tools

Research boundary

Experimental research software. QSA measurements are workload-specific, and the public record does not establish a universal quantum advantage or production-LLM replacement claim.

Connected work