top of page

Overview

 

My research lies at the intersection of quantum many-body physics, tensor network algorithms, quantum computation, and machine learning. I explore both foundational questions in quantum theory and practical computational tools for simulating and optimizing quantum systems. This includes developing simulation frameworks, scalable algorithms, and hybrid methods tailored for the emerging era of quantum technologies.

Tensor Networks for Quantum Many-Body Systems

I develop and apply advanced tensor network algorithms such as MPS, PEPS, Flexible-PEPS, and DMRG to study highly entangled and strongly correlated systems. My contributions include:

​

  • Designing the Flexible-PEPS algorithm to simulate quantum systems on irregular, non-Euclidean geometries (e.g., hyperbolic lattices).

  • Simulating frustrated spin systems, quantum liquids, and topological phases.

  • Using tensor networks to extract entanglement structures, correlation profiles, and phase transitions.

  • These techniques push beyond traditional MPS approaches and open new directions in the study of exotic quantum phases.

Quantum Circuit Simulation & Benchmarking

  • I built a tensor network-based quantum circuit simulator that has demonstrated superior performance to IBM’s QPUs, Qiskit simulators, and CUDA-Q on several benchmarks.

  • Simulates deep quantum circuits in VQE, QAOA, and quantum error correction.

  • Benchmarks real hardware using structured entanglement contraction.

  • Supports both CPU and GPU execution with realistic noise modeling.

  • This tool contributes to validating experimental quantum processors and supporting algorithm design in the NISQ regime.

Hybrid Algorithms and Quantum Machine Learning

  • My work connects tensor network simulations with hybrid quantum-classical algorithms for practical applications in physics, chemistry, and optimization:

  • Integrated Flexible-PEPS with VQE and QAOA frameworks.

  • Modeled NP-hard combinatorial problems using tensor-based encodings.

  • These methods enable high-accuracy modeling of quantum systems and real-world optimization problems within current hardware limitations.

AI for Physics

​

  • Constructed tensorized neural networks in PyTorch for creating memory efficient Deep Neural Networks.

  • Investigated integer-based optimization techniques as an alternative to standard gradient-based training in AI models.

  • ​

  • Future direction:

  • Develop Neural Quantum States (NQS) to model systems with volume-law entanglement.

  • Investigate CNNs and RBMs for learning quantum correlations.

  • Explore hybrid NQS-tensor network architectures for scalable quantum state representation.

  • This direction aims to bridge the gap between AI expressivity and quantum many-body complexity.

bottom of page