用量子电路提升大模型性能，仅用6000参数在真实量子机上实现1.4%困惑度降低。
Quantum-enhanced Large Language Models on Quantum Hardware via Cayley Unitary Adapters
- 用凯利参数化量子模块嵌入冻结的LLM投影层，实现可训练量子组件。
- 在156量子比特处理器上，使Llama 3.1 8B困惑度降低1.4%，仅增6000参数。
- 发现噪声-表达力相变现象，为未来更大规模量子优势提供路径指引。
大型语言模型（LLMs）已重塑人工智能，但经典架构存在根本限制：每个可训练参数都需随模型规模增长的古典内存。量子计算提供了质的差异路径，但对实际规模模型在真实硬件上的演示仍属空白。本文展示，将凯利参数化酉适配器——插入预训练LLM冻结投影层的量子电路模块——在156量子比特的IBM Quantum System Two超导处理器上运行，使广泛使用的80亿参数模型Llama 3.1 8B的困惑度降低1.4%，仅增加6000个额外参数，并完成了端到端推理验证。对可处理的SmolLM2（135M参数）进行系统研究，揭示了酉模块维度与困惑度单调提升的关系，恢复了83%压缩导致的性能退化，并正确回答了两个经典基线均失败的问题；同时识别出清晰的噪声-表达力相变，指明了未来在更大量子比特规模下实现量子优势的具体路径。
原文摘要 · Abstract (English)
Large language models (LLMs) have transformed artificial intelligence, yet classical architectures impose a fundamental constraint: every trainable parameter demands classical memory that scales unfavourably with model size. Quantum computing offers a qualitatively different pathway, but practical demonstrations on real hardware have remained elusive for models of practical relevance. Here we show that Cayley-parameterised unitary adapters -- quantum circuit blocks inserted into the frozen projection layers of pre-trained LLMs and executed on a 156-qubit IBM Quantum System Two superconducting processor -- improve the perplexity of Llama 3.1 8B, an 8-billion-parameter model in widespread use, by 1.4% with only 6,000 additional parameters and end-to-end inference validated on real Quantum Processing Unit (QPU). A systematic study on SmolLM2 (135M parameters), chosen for its tractability, reveals monotonically improving perplexity with unitary block dimension, 83% recovery of compression-induced degradation, and correct answers to questions that both classical baselines fail -- with a sharp noise-expressivity phase transition identifying the concrete path to quantum utility at larger qubit scales.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。