Research map / Learning theory and algorithms / Kernels and regression theory
Symmetric Pruning in Quantum Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add SP1 (prune qubit wires on identity part) (ablation) | HEA with symmetric pruning · TFIM | parameters at critical point: 800 -> 300 | Helped |
| add SP1 (prune qubit wires on identity part) (ablation) | HEA with symmetric pruning · MaxCut | parameters at critical point: 1000 -> 300 | Helped |
| add SP2 (prune structure) after SP1 (ablation) | HEA with symmetric pruning · TFIM | parameters at critical point: 300 -> 120 | Helped |
| add SP2 (prune structure) after SP1 (ablation) | HEA with symmetric pruning · MaxCut | parameters at critical point: 300 -> 150 | Helped |
| add SP3 (correlate symmetric parameters) after SP2 (ablation) | HEA with symmetric pruning · TFIM | parameters at critical point: 120 -> 50 | Helped |
| add SP3 (correlate symmetric parameters) after SP2 (ablation) | HEA with symmetric pruning · MaxCut | parameters at critical point: 150 -> 100 | Helped |
| full symmetric pruning SP0 -> SP3 (ablation) | HEA with symmetric pruning · TFIM | EQNTK value Q_S: 20 | Helped |
| full symmetric pruning SP0 -> SP3 (ablation) | HEA with symmetric pruning · MaxCut | EQNTK value Q_S: 14 | Helped |
| symmetric pruning SP1-SP3 vs unpruned, LK>=300 (ablation) | HEA with symmetric pruning · TFIM | iterations to epsilon-convergence: up to 6000 steps | Helped |
| symmetric pruning SP1-SP3 vs unpruned, LK>=300 (ablation) | HEA with symmetric pruning · MaxCut | iterations to epsilon-convergence: up to 5000 steps | Helped |
| increase EQNTK from SP1 to SP3 at LK~450 (ablation) | HEA with symmetric pruning · MaxCut | iterations T(epsilon): Q 1:2:4, T 4:2:1 | Helped |
| increase number of parameters within the same ansatz design | HEA with symmetric pruning · TFIM and MaxCut | convergence speed | Helped |
Closest papers by meaning
- ATP: Adaptive Threshold Pruning for Efficient Data Encoding in Quantum Neural Networks (CVPR 2025 · similarity 0.88)
- Exponential Hardness of Optimization from the Locality in Quantum Neural Networks (AAAI 2024 · similarity 0.87)
- Exploring channel distinguishability in local neighborhoods of the model space in quantum neural networks (ICLR 2025 · similarity 0.86)
- Quantum Deep Equilibrium Models (NeurIPS 2024 · similarity 0.86)
- One More Time: Revisiting Neural Quantum States from a Reinforcement Learning Perspective (NeurIPS 2026 · similarity 0.86)
- Diagonal Symmetrization of Neural Network Solvers for the Many-Electron Schrödinger Equation (ICML 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.