# Neural QAOA2: Differentiable Joint Graph Partitioning and Parameter Initialization for Quantum Combinatorial Optimization

arXiv: https://arxiv.org/abs/2605.13072 · HTML: https://mlsanity.com/paper/2605.13072/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 61 references that are in the map.
- First posted on arXiv 13 May 2026.

## Builds on (most cited first, 27 of 27)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables](https://mlsanity.com/paper/1611.00712/) (2016)
- [Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning](https://mlsanity.com/paper/1801.07606/) (2018)
- [How Attentive are Graph Attention Networks?](https://mlsanity.com/paper/2105.14491/) (2021)
- [Hierarchical Graph Representation Learning with Differentiable Pooling](https://mlsanity.com/paper/1806.08804/) (2018)
- [OptNet: Differentiable Optimization as a Layer in Neural Networks](https://mlsanity.com/paper/1703.00443/) (2017)
- [Learning Combinatorial Optimization Algorithms over Graphs](https://mlsanity.com/paper/1704.01665/) (2017)
- [Differentiable Convex Optimization Layers](https://mlsanity.com/paper/1910.12430/) (2019)
- [Ising formulations of many NP problems](https://mlsanity.com/paper/1302.5843/) (2013)
- [Learning Latent Permutations with Gumbel-Sinkhorn Networks](https://mlsanity.com/paper/1802.08665/) (2018)
- [Graph Clustering with Graph Neural Networks](https://mlsanity.com/paper/2006.16904/) (2020)
- [Large Language Models as Evolutionary Optimizers](https://mlsanity.com/paper/2310.19046/) (2023)
- [Combinatorial Optimization with Physics-Inspired Graph Neural Networks](https://mlsanity.com/paper/2107.01188/) (2021)
- [LLMOPT: Learning to Define and Solve General Optimization Problems from Scratch](https://mlsanity.com/paper/2410.13213/) (2024)
- [Estimating or Propagating Gradients Through Stochastic Neurons](https://mlsanity.com/paper/1305.2982/) (2013)
- [Bridging Discrete and Backpropagation: Straight-Through and Beyond](https://mlsanity.com/paper/2304.08612/) (2023)
- [How Good Is Neural Combinatorial Optimization? A Systematic Evaluation on the Traveling Salesman Problem](https://mlsanity.com/paper/2209.10913/) (2022)
- [Controlling Continuous Relaxation for Combinatorial Optimization](https://mlsanity.com/paper/2309.16965/) (2023)
- [LLM-Driven Instance-Specific Heuristic Generation and Selection](https://mlsanity.com/paper/2506.00490/) (2025)
- [Graph neural network initialisation of quantum approximate optimisation](https://mlsanity.com/paper/2111.03016/) (2021)
- [Optimization by Parallel Quasi-Quantum Annealing with Gradient-Based Sampling](https://mlsanity.com/paper/2409.02135/) (2024)
- [Free-Energy Machine for Combinatorial Optimization](https://mlsanity.com/paper/2412.09285/) (2024)
- [MaxCutPool: differentiable feature-aware Maxcut for pooling in graph neural networks](https://mlsanity.com/paper/2409.05100/) (2024)
- [MG-Net: Learn to Customize QAOA with Circuit Depth Awareness](https://mlsanity.com/paper/2409.18692/) (2024)
- [Solving Quadratic Unconstrained Binary Optimization with divide-and-conquer and quantum algorithms](https://mlsanity.com/paper/2101.07813/) (2021)
- [MLQAOA: Graph Learning Accelerated Hybrid Quantum-Classical Multilevel QAOA](https://mlsanity.com/paper/2404.14399/) (2024)
- [Cascaded Large-Scale TSP Solving with Unified Neural Guidance: Bridging Local and Population-based Search](https://mlsanity.com/paper/2501.14285/) (2025)

## Built on by (most cited first, 1 of 1)

- [Beyond QAOA: A Review of AI and Quantum Computing for Adaptive Combinatorial Optimization](https://mlsanity.com/paper/2610.11759/) (2026)

## Abstract (arXiv)

The quantum approximate optimization algorithm (QAOA) holds promise for combinatorial optimization but is constrained by limited qubits. While divide-and-conquer frameworks like QAOA2 address scalability by partitioning graphs into subgraphs, existing methods suffer from two fundamental limitations: i) misalignment between heuristic partitioning metrics and quantum optimization goals, and ii) topology-blind parameter initialization that leads to optimization cold starts. To bridge these gaps, we propose Neural QAOA2, an end-to-end differentiable framework that jointly generates graph partitions and initial parameters. By integrating a generative evaluative network (GEN), our method utilizes a differentiable quantum evaluator as a high-fidelity performance surrogate to provide direct gradient guidance, enabling the joint generator to learn the intrinsic mapping from graph topology to high-quality partition and parameter configurations. Extensive experiments on 183 QUBO, Ising, and MaxCut instances (21 to 1000 variables) demonstrate that our gradient-driven approach broadly outperforms heuristic baselines, ranking first on 101 instances. It exhibits zero-shot generalization across out-of-distribution graph topologies and scales.