Skip to content

Watching Quantum Models Think: Hilbert-Space Interpretability in Quantum Transformer Blocks

Sep 2026 · 0 citations · 26 references
Physics Computer Science

TL;DR

These proof-of-concept results, obtained on small synthetic tasks, suggest that the physics of quantum computation can provide intrinsic interpretability signals with no direct classical counterpart, motivating study of whether this advantage persists at scale.

Abstract

Deep learning models are powerful but opaque. As quantum machine learning matures, the field faces a defining choice: build quantum models that are equally opaque, or exploit the mathematical structure of quantum mechanics to make them inherently interpretable. We show that the latter is possible. By tracking quantum mutual information~(MI), entanglement entropy, and state fidelity through the layers of a Quantum Transformer Block (\qtb{}), a fully-coherent variational circuit with quantum analogues of both attention and feedforward, we gain direct insight into how the model processes information: which tokens it attends to, when correlations form, and why predictions fail. On four tasks with known dependency structure we show that (i)~learned MI matrices align with ground-truth task structure (AUC$\,{=}\,0.69$ on lookup), (ii)~disabling entangling gates collapses accuracy from 100\% to 15\% while MI$\to 0$, proving entanglement is the mechanism, (iii)~accuracy and MI co-evolve during training ($\rho\,{=}\,0.92$ on lookup), and (iv)~per-sample MI predicts prediction correctness on the conditional task with ROC AUC$\,{=}\,0.84$. All results are validated on IBM Quantum hardware (ibm\_kingston, Heron~r2): the circuit's reasoning process, from product state through structured entanglement, is directly observable on a superconducting processor. These proof-of-concept results, obtained on small synthetic tasks, suggest that the physics of quantum computation can provide intrinsic interpretability signals with no direct classical counterpart, motivating study of whether this advantage persists at scale.

View source

Similar papers

#quantum computing Preprint Oct 2026

Can the Quantum Fisher Effective Dimension Judge a Hybrid Quantum Neural Network?

Hybrid quantum neural networks (QNNs) pair a classical feature extractor with a variational quantum circuit, and the effective dimension of the circuit's quantum Fisher information matrix (QFIM) is widely used to assess such models. This scalar is often interpreted as an indicator of trainability or accuracy, although...

M. Mia, A. Khandakar, M. Ayari · 0 citations
Preprint Sep 2026

Quantum Representation Learning Beyond Pairwise Fidelity

It is shown that a transition-probability-only learning interface can possess exact continuous blind directions in certain quantum-state families, and multistate relational observables are established as measurable, trainable, and physically consequential signals for quantum representation learning beyond pairwise fide...

Jun-Peng Hou, Chang-Bin Lu · 0 citations
#artificial intelligence Preprint Sep 2026

QiT: Quantum-Inspired Transformer for Visual Recognition Task

QT is introduced, a Quantum-inspired Transformer for vision tasks with three components: angle-inspired encoding that maps image tokens to learned trigonometric Hilbert-space features analogous to quantum rotation-based state encoding; self-attention over these periodic features, inducing a classical cosine kernel appr...

B. Patro, V. Agneeswaran · 0 citations
Preprint Oct 2026

Who can sample forever? Shot noise effects in quantum linear regression

Quantum fidelity kernels and quantum extreme learning machines share the same basic architecture: a fixed quantum device generates features that are combined through a classically trained linear model. We refer to this broad non-variational setting as quantum linear regression. A defining aspect of these models is that...

G. Lo Monaco, S. Lorenzo, A. Ferraro et al. · 1 citation
#machine learning Preprint Oct 2026

Generalization of Transformer-Based Neural Quantum States via In-Context Learning

Neural quantum states based on modern deep learning architectures have emerged as powerful representations for quantum many-body systems. In particular, Transformer-based neural quantum states provide expressive models capable of capturing long-range correlations, and their empirical generalization performance has rece...

Zhen Qin, Qing-Da Qu, A. Hero · 0 citations

Related blog posts

MIT News · Artificial Intelligence Nov 17, 2026

Learning the language of machine learning

By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.

MIT News · Artificial Intelligence Oct 28, 2026

Making machines that can listen

Professor Paris Smaragdis develops computers with superhuman hearing for applications like speech recognition, music understanding, and environmental monitoring.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.