Quantum-informed machine learning for predicting spatiotemporal chaos

ArticleMaida Wang, Xiao Xue, Mingyang Gao, Peter V. CoveneyResearch

Researchers from University College London and the University of Amsterdam have introduced QIML, a hybrid framework combining quantum generative models with classical autoregressive predictors. The method stabilises long-term forecasting of high-dimensional chaotic systems, offering practical advantages over purely classical baselines.

Key points

  1. QIML integrates a one-time, offline-trained quantum generative model with a classical autoregressive predictor for spatiotemporal field generation.
  2. The quantum component learns a 'Q-Prior' representing fine-scale interactions and small-scale dynamics.
  3. Evaluations across the Kuramoto-Sivashinsky equation, 2D Kolmogorov flow, and 3D turbulent channel flow showed improvements of up to 17.25% in predictive distribution accuracy and up to 29.36% in full-spectrum fidelity.
  4. The Q-Prior for turbulent channel inflow was trained on a superconducting quantum processor and proved essential for preventing prediction instability.
  5. QIML compresses multi-megabyte datasets into a kilobyte-scale prior, bypassing recurring quantum-classical communication overhead during inference.

Why it matters

Classical neural operators and deep learning architectures struggle with long-term stability in chaotic and turbulent regimes due to error divergence. QIML illustrates a practical path for current noisy intermediate-scale quantum (NISQ) hardware to enhance complex macroscopic physical modeling without relying on full fault tolerance.

Analysis

A major technical obstacle in quantum-enhanced machine learning has been the latency and measurement bottleneck caused by repeated data exchanges between classical systems and quantum hardware. By restricting the quantum component to a one-time offline generation of the invariant Q-Prior, the authors eliminate runtime roundtrips while preserving the quantum circuit's capacity to represent nonlocal Hilbert space distributions. This offers both a significant memory compression benefit and superior stability compared with standard PDE solvers.

Mentioned links

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Article:NVIDIA et al.

    Pretraining LLMs in 4-Bit: NVIDIA Introduces NVFP4 Training Methodology

    A research team from NVIDIA has presented a method for stable and accurate pretraining of Large Language Models using the NVFP4 4-bit floating-point format. To evaluate the approach, they trained a 12-billion-parameter model on 10 trillion tokens—reported as the longest publicly documented 4-bit training run to date. The model matched the training loss and downstream task performance of an FP8 baseline.

    AI & LLMs· Research

  • Article:Maximilian Beck, Korbinian Pöppel, Markus Spanring, Andreas Auer, Oleksandra Prudnikova, Michael Kopp, Günter Klambauer, Johannes Brandstetter, Sepp Hochreiter

    xLSTM: Extended Long Short-Term Memory

    A research team led by Sepp Hochreiter introduces xLSTM, an extension of the original Long Short-Term Memory architecture designed for large-scale language modelling. The approach addresses historical scaling limits of LSTMs and demonstrates competitive performance against state-of-the-art Transformers and State Space Models.

    AI & LLMs· Research

  • Link:Zhe Liu, Jinghua Hou, Yuxiang Lu, et al.

    StreamPI: Streaming Multimodal Temporal Modeling for Vision-Language-Action Models

    Researchers from The University of Hong Kong and ACE Robotics introduce StreamPI, a framework that equips single-frame Vision-Language-Action (VLA) models such as π0.5 with temporal reasoning without adding new parameters. By bundling images and instructions into atomic temporal units and using random-interval training, StreamPI improves precision, memory, and robustness in real-robot manipulation.

    AI & LLMs· Research

  • Article:Peiyan Li, Yuze Zhu, Yixiang Chen, Qisen Ma, Yuan Xu, Jiabing Yang, He Guan, Yan Huang, Hongtao Wu, Xiao Ma, Tao Kong, Liang Wang, Tieniu Tan

    BridgeVLA++: A Memory-Augmented 3D Vision-Language-Action Framework for Robotics

    Researchers introduce BridgeVLA++, an upgraded vision-language-action framework designed for 3D robotic manipulation. It introduces a unified spatio-temporal memory architecture into pre-trained vision-language models, enabling persistent spatial reasoning and temporal history tracking without losing data efficiency.

    AI & LLMs· Research

  • Article:Yoel Shoshan, Moshiko Raboh, Michal Ozery-Flato, et al.

    MAMMAL: Multi-Modal Foundation Model for Biomedical Discovery

    Researchers from IBM and collaborating institutions introduce MAMMAL, a cross-modal foundation model tailored to drug discovery. The architecture unifies protein and antibody sequences, small molecules, and gene expression profiles into a sequence-based framework, pre-trained on two billion samples.

    AI & LLMs· Research

  • Link:Discovery Loop

    Discovery Loop: Automating the Scientific Discovery Process

    Jeff Dean, Sanjay Ghemawat, Quoc Le, and Oriol Vinyals have introduced Discovery Loop, an effort to automate scientific experimental loops using frontier AI and large-scale computing infrastructure, starting with ML research.

    AI & LLMs· Announcement

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.