Update README.md
Browse files
    	
        README.md
    CHANGED
    
    | @@ -1,4 +1,90 @@ | |
| 1 | 
             
            ---
         | 
| 2 | 
             
            license: apache-2.0
         | 
| 3 | 
             
            pipeline_tag: tabular-regression
         | 
| 4 | 
            -
             | 
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | |
|  | 
|  | |
| 1 | 
             
            ---
         | 
| 2 | 
             
            license: apache-2.0
         | 
| 3 | 
             
            pipeline_tag: tabular-regression
         | 
| 4 | 
            +
            datasets:
         | 
| 5 | 
            +
            - Allanatrix/QST
         | 
| 6 | 
            +
            tags:
         | 
| 7 | 
            +
            - Physics
         | 
| 8 | 
            +
            ---
         | 
| 9 | 
            +
            # NexaQST: Quantum State Tomography with Physics-Informed Neural Networks
         | 
| 10 | 
            +
             | 
| 11 | 
            +
            **NexaQST** is a Quantum State Tomography model built using a Physics-Informed Neural Network (PINN) trained on synthetic 2-qubit experiments. This model leverages quantum mechanical priors such as positivity, Hermiticity, and trace constraints to ensure physically plausible reconstructions of quantum states.
         | 
| 12 | 
            +
             | 
| 13 | 
            +
            ---
         | 
| 14 | 
            +
             | 
| 15 | 
            +
            ## Model Overview
         | 
| 16 | 
            +
             | 
| 17 | 
            +
            - **Task**: Reconstruct quantum states (density matrices) from tomographic measurement traces
         | 
| 18 | 
            +
            - **System**: Simulated 2-qubit experiments
         | 
| 19 | 
            +
            - **Architecture**: Physics-Informed Neural Network (PINN)
         | 
| 20 | 
            +
            - **Constraints Embedded**:
         | 
| 21 | 
            +
              - Positivity
         | 
| 22 | 
            +
              - Hermiticity
         | 
| 23 | 
            +
              - Trace normalization (Tr(Ο) = 1)
         | 
| 24 | 
            +
             | 
| 25 | 
            +
            ---
         | 
| 26 | 
            +
             | 
| 27 | 
            +
            ## Dataset Generation
         | 
| 28 | 
            +
             | 
| 29 | 
            +
            - **Simulation Toolkit**: [`QuTiP`](https://qutip.org/)
         | 
| 30 | 
            +
            - **Process**:
         | 
| 31 | 
            +
              1. Generated full 2-qubit tomography experiments via simulation
         | 
| 32 | 
            +
              2. Extracted and structured measurement traces
         | 
| 33 | 
            +
              3. Derived backward synthetic traces from known density matrices
         | 
| 34 | 
            +
              4. Created supervised training pairs: `(trace β density matrix)`
         | 
| 35 | 
            +
              5. Enforced physical constraints directly inside the model during training
         | 
| 36 | 
            +
             | 
| 37 | 
            +
            ---
         | 
| 38 | 
            +
             | 
| 39 | 
            +
            ## Model Input/Output
         | 
| 40 | 
            +
             | 
| 41 | 
            +
            - **Input**: Vector of tomographic measurement traces
         | 
| 42 | 
            +
            - **Output**: Reconstructed density matrix (complex-valued 4x4 for 2 qubits)
         | 
| 43 | 
            +
             | 
| 44 | 
            +
            ---
         | 
| 45 | 
            +
             | 
| 46 | 
            +
            ## Example Usage
         | 
| 47 | 
            +
             | 
| 48 | 
            +
            ```python
         | 
| 49 | 
            +
            # Pseudocode for usage
         | 
| 50 | 
            +
             | 
| 51 | 
            +
            from model import NexaQSTModel
         | 
| 52 | 
            +
            import torch
         | 
| 53 | 
            +
             | 
| 54 | 
            +
            model = NexaQSTModel()
         | 
| 55 | 
            +
            model.load_state_dict(torch.load("nexaqst_model.pt"))
         | 
| 56 | 
            +
            model.eval()
         | 
| 57 | 
            +
             | 
| 58 | 
            +
            traces = load_measurement_vector("qst_trace.npy")  # shape: (N,)
         | 
| 59 | 
            +
            rho_pred = model.predict_density_matrix(traces)
         | 
| 60 | 
            +
             | 
| 61 | 
            +
            # rho_pred is a 4x4 complex-valued matrix satisfying physical constraints
         | 
| 62 | 
            +
            ````
         | 
| 63 | 
            +
             | 
| 64 | 
            +
            ---
         | 
| 65 | 
            +
             | 
| 66 | 
            +
            ## Applications
         | 
| 67 | 
            +
             | 
| 68 | 
            +
            * Quantum error correction diagnostics
         | 
| 69 | 
            +
            * Quantum system identification
         | 
| 70 | 
            +
            * Educational simulation of QST techniques
         | 
| 71 | 
            +
            * Physically consistent state estimation for quantum simulations
         | 
| 72 | 
            +
             | 
| 73 | 
            +
            ---
         | 
| 74 | 
            +
             | 
| 75 | 
            +
            ## Licensing & Citation
         | 
| 76 | 
            +
             | 
| 77 | 
            +
            * If used in publications or products, please credit the Nexa Scientific Suite.
         | 
| 78 | 
            +
             | 
| 79 | 
            +
            ---
         | 
| 80 | 
            +
             | 
| 81 | 
            +
            ## π Related Tools
         | 
| 82 | 
            +
             | 
| 83 | 
            +
            * [π Nexa Data Studio](https://huggingface.co/spaces/Allanatrix/NexaDataStudio)
         | 
| 84 | 
            +
            * [π¬ Nexa R\&D Hub](https://huggingface.co/spaces/Allanatrix/NexaR&D)
         | 
| 85 | 
            +
            * [π§  Nexa MOE Models](https://huggingface.co/collections/Allanatrix/nexa-models)
         | 
| 86 | 
            +
            * [π Nexa Hub (Main Portal)](https://huggingface.co/spaces/Allanatrix/NexaHub)
         | 
| 87 | 
            +
             | 
| 88 | 
            +
            ---
         | 
| 89 | 
            +
             | 
| 90 | 
            +
            *Created by [Allan](https://huggingface.co/Allanatrix), independent quantum systems architect and ML researcher. Part of the Nexa scientific computing ecosystem.*
         | 
