Quantum-Inspired Fine-Tuning for Few-Shot AIGC Detection via Phase-Structured Reparameterization

TL;DR

This paper introduces Q-LoRA, a quantum-inspired fine-tuning method that enhances few-shot AI-generated content detection, and proposes H-LoRA, a classical variant that retains similar benefits with lower computational overhead.

Contribution

The paper presents Q-LoRA, a novel quantum-inspired fine-tuning scheme for few-shot learning, and introduces H-LoRA, a classical alternative that achieves comparable performance with reduced cost.

Findings

Q-LoRA outperforms standard LoRA in few-shot AIGC detection.

H-LoRA achieves similar accuracy to Q-LoRA with lower computational overhead.

Both methods improve accuracy by over 5% compared to standard LoRA.

Abstract

Recent studies show that quantum neural networks (QNNs) generalize well in few-shot regimes. To extend this advantage to large-scale tasks, we propose Q-LoRA, a quantum-enhanced fine-tuning scheme that integrates lightweight QNNs into the low-rank adaptation (LoRA) adapter. Applied to AI-generated content (AIGC) detection, Q-LoRA consistently outperforms standard LoRA under few-shot settings. We analyze the source of this improvement and identify two possible structural inductive biases from QNNs: (i) phase-aware representations, which encode richer information across orthogonal amplitude-phase components, and (ii) norm-constrained transformations, which stabilize optimization via inherent orthogonality. However, Q-LoRA incurs non-trivial overhead due to quantum simulation. Motivated by our analysis, we further introduce H-LoRA, a fully classical variant that applies the Hilbert transform within the LoRA adapter to retain similar phase structure and constraints. Experiments on few-shot AIGC detection show that both Q-LoRA and H-LoRA outperform standard LoRA by over 5% accuracy, with H-LoRA achieving comparable accuracy at significantly lower cost in this task.