Quantum Language Model Uses Boltzmann Machine for Self-Attention

Researchers have developed a novel language model, the FC-QBM-LM, which utilizes a fully connected quantum Boltzmann machine (FC-QBM) as its core engine. This model incorporates a self-attention mechanism that performs unitary evolution on semantic quantum states, implicitly handling positional encoding and self-attention computation without discrete matrix operations. Experiments with an 8-qubit FC-QBM-LM trained on Tang poetry demonstrated favorable convergence, strong noise robustness, and preliminary capabilities in associative thinking, content generation, and word ordering. AI

IMPACT Introduces a novel quantum approach to language modeling, potentially offering new avenues for efficient computation and robustness.

RANK_REASON Academic paper describing a novel model architecture and its experimental validation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• FC-QBM-LM
• Fully Connected Quantum Boltzmann Machine
• Quantum Boltzmann Machine
• Tang poetry
• X-axis rotation gates
• Y-axis rotation gates

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →