- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
This paper proposes a single-layer spiking graph neural network that replaces conventional spatial layer stacking with temporal evolution of spiking neuron dynamics. The input current of each LIF neuron is decoupled into an external driving current and an internal interaction current, where the latter consists of a graph-attention current and a Hopfield associative-memory current. The authors construct a global energy functional with membrane-recovery, internal-aggregation, and external-driving terms, and provide theoretical analyses showing monotonic energy decrease under soft-threshold continuous dynamics and hard-threshold discrete state flips. Experiments on eight node-classification benchmarks show strong performance among spiking GNNs, competitive results against continuous GNNs, resistance to over-smoothing under long temporal evolution, and favorable theoretical computational cost. Overall, the paper presents a biologically motivated and theoretically grounded alternative to hierarchical spiking GNNs.
Critical strengths
-
The temporal-evolution paradigm is a clean and biologically plausible alternative to spatial stacking in spiking GNNs.
-
The decoupling of external driving and internal graph-attention/Hopfield currents is well motivated and leads to an interpretable energy-based formulation.
-
The paper provides convergence/stability analyses for both soft-threshold and hard-threshold firing, which is stronger than many empirical spiking GNN papers.
Critical weaknesses
-
The energy functional is used for theoretical analysis, but it is not clear whether it participates in the training objective or only serves as an analytical tool. This should be stated explicitly.
-
The ablation study focuses on current components but does not sufficiently analyze key hyperparameters.
Questions for authors
-
How often does the condition hold empirically for trained models on the eight datasets?
-
Does the energy-descent guarantee still hold under synchronous hard-threshold updates used in Algorithm 1, or only under asynchronous updates?
-
What are the training-time memory and runtime costs compared with SpikeGCL, SpikeNet, and at least one ODE-based GNN?
-
How does the method perform on heterophilous graphs and on tasks beyond node classification?
2 comments
rebuttal (2/2)
Q5. How do training time and GPU memory usage compare with SpikeGCL, SpikeNet, and ODE models?
Figure 6 already reports training cost comparisons with SpikeGCL and SpikeNet. Following your suggestion, we retained the original experimental configuration and added the ODE model GREAD. The results below were obtained on the Computers dataset using an RTX 3090 with , and the timing statistics use 5 random seeds.
| Method | Time per epoch (ms) | Peak GPU memory during training (MiB) |
|---|---|---|
| Our method | 83.98 ± 0.53 | 773.45 |
| SpikeGCL | 102.99 ± 2.80 | 721.81 |
| SpikeNet | 219.02 ± 9.33 | 1045.92 |
| GREAD | 78.57 ± 0.32 | 3902.64 |
Under these configurations, our time per epoch is reduced by approximately 18.5% and 61.7% compared with SpikeGCL and SpikeNet, respectively. GREAD is slightly faster per epoch, but our peak GPU memory usage during the procedure is approximately 80.2% lower. In addition, at in Figure 6, the peak GPU memory usage of our method and SpikeGCL is 848.30 and 1665.74 MiB, respectively, demonstrating the memory advantage with longer temporal unfolding.
Q6. How does the method perform on heterophilous graphs and tasks beyond node classification?
Heterophilous graphs. Our local interaction mechanism is more naturally suited to homophilous graphs because it tends to strengthen related states between adjacent nodes. On strongly heterophilous graphs, one-hop neighbors may contain information inconsistent with the central node's label, so relying solely on local neighborhood interactions may limit the model's performance.
However, the external current design in our framework is highly flexible and can explicitly incorporate multi-hop structural information.
Here, denotes the encoded node features, is the graph propagation matrix, is the learnable mapping for each hop, and is the maximum number of aggregation hops. corresponds to the original external current, while additionally incorporate one-hop and two-hop information, respectively. We conducted experiments on Chameleon and Squirrel using the 10 publicly available fixed splits provided by Geom-GCN [1]. The test accuracies are shown below.
| Method | Chameleon (%) | Squirrel (%) |
|---|---|---|
| SpikeNet | 60.68 ± 1.43 | 40.49 ± 1.03 |
| SpikeGCL | 75.29 ± 1.42 | 59.61 ± 1.22 |
| MSG | 42.02 ± 2.95 | 27.48 ± 1.31 |
| Ours() | 54.82 ± 1.93 | 39.00 ± 1.07 |
| Ours() | 66.58 ± 1.26 | 49.84 ± 1.20 |
| Ours() | 67.43 ± 1.62 | 50.71 ± 1.56 |
Compared withK=0 , both and lead to clear accuracy improvements on Chameleon and Squirrel. These results demonstrate that incorporating multi-hop structural information into the external current can improve performance on heterophilic graphs. Although our method does not yet achieve the best performance under this setting, the results validate the extensibility of our external-current design to heterophilic scenarios. A more systematic design of heterophily-aware external currents will be investigated in future work.
Graph-level tasks. Our method can be naturally extended to graph-level tasks. Without modifying the underlying dynamics, we only need to replace the node-level readout layer with a graph-level readout operation.
We conducted additional experiments on two graph-level regression datasets, ESOL and FreeSolv. Each dataset was split into training, validation, and test sets using an 8:1:1 ratio. The evaluation metric is the test RMSE, and all results are reported as the mean ± standard deviation over five random seeds.
| Method | ESOL RMSE ↓ | FreeSolv RMSE ↓ |
|---|---|---|
| GCN | 0.801 ± 0.039 | 1.257 ± 0.051 |
| GAT | 0.668 ± 0.015 | 1.085 ± 0.069 |
| SpikeGCL | 0.908 ± 0.054 | 1.616 ± 0.080 |
| SpikeGraphormer | 0.695 ± 0.049 | 1.004 ± 0.161 |
| Ours | 0.678 ± 0.025 | 0.902 ± 0.037 |
Our method achieves the best performance on FreeSolv and competitive performance on ESOL, demonstrating that the proposed dynamics can also be applied to graph-level prediction tasks.
It should be noted that graph-level tasks are fundamentally different from node-classification tasks in terms of the energy formulation. For node classification, the energy function is constructed over the entire graph dataset. In contrast, for graph-level tasks, an energy function is constructed independently for each graph sample.
[1] Geom-GCN: Geometric Graph Convolutional Networks
rebuttal(1/2)
We thank the reviewer for the positive assessment of our temporal evolution architecture, current decomposition, and energy analysis. In response to your questions, we have added relevant explanations and experimental results and respond point by point below.
Q1. Is the energy function involved in the training objective?
The parameters in the energy function are continuously updated during training. In fact, the energy function is not used as an additional training loss, but it can describe the state of the network's evolution. The energy function unifies membrane recovery, graph attention, Hopfield memory, and sustained external drive within the same dynamical framework and is used to analyze its stability. We will explain this in the paper.
Q2. How sensitive is the method to key hyperparameters?
We added sensitivity experiments on PubMed and CS for the number of memories M, the time constant τ, and the step size dt. The default setting is M=256,τ=2,dt=1. Only one parameter was changed at a time, and each configuration was trained with 3 seeds for 200 epochs. The table below reports validation accuracy (%), with the sweep columns reporting means in the order given in their headings.
Although different hyperparameter choices lead to small performance variations, none of the individual hyperparameters causes large fluctuations in model performance within the tested range. These results demonstrate that our model remains relatively stable over reasonable ranges of hyperparameter values.
Q3. Does the condition σ2(HIext)>0 hold in the trained models?
The minimum second singular values of the trained models on seven datasets are shown below.
All values are clearly greater than zero, indicating that the centered external currents of the evaluated models retain at least two linearly independent directions of variation across nodes, supporting the nondegeneracy condition required by Theorem 1.
Q4. Do synchronous hard-threshold updates have an energy descent guarantee?
The current proof of hard-threshold energy descent applies to asynchronous updates. For the synchronous updates in Algorithm 1, we have not yet established a corresponding rigorous descent guarantee. We will clarify this scope in the revised manuscript and further investigate sufficient conditions for descent and theoretical analysis of synchronous updates in future work.