---
title: Dynamic Graph Neural Networks
url: https://www.emergentmind.com/topics/dynamic-graph-neural-networks-dgnns
type: topic
---

# Dynamic Graph Neural Networks

Dynamic Graph Neural Networks (DGNNs) are a class of neural architectures that combine structural representation learning with temporal modeling for graphs whose topology and attributes evolve over time. DGNNs have established themselves as the state-of-the-art for supervised and self-supervised learning on dynamic graphs, addressing core challenges in domains such as social interaction modeling, recommender systems, financial transaction networks, biological systems, and physical process forecasting. Central to DGNNs is the integration of topological (spatial) information and temporal dependencies, yielding time-aware node and edge embeddings suitable for prediction tasks inherently tied to temporal evolution.

## 1. Formal Foundations and Model Taxonomy

A dynamic graph is a system in which the vertex set and edge set change over time, with associated node and edge attributes possibly varying as well. Two common representations are:

- **Discrete-time dynamic graphs (DTDG):** Sequences of graph snapshots $\{G^1, \ldots, G^T\}$, where each $G^t=(V^t, E^t, X^t)$.  
- **Continuous-time dynamic graphs (CTDG):** Streams of timestamped interaction events $\{\left(u_i, v_i, t_i\right)\}$.

The goal of DGNNs is to learn functions $f_\theta$ that encode both instantaneous structural information and the temporal evolution, producing embeddings $h_{i}^{t}$ per node over time for downstream prediction tasks. The model taxonomy can be stratified as follows [2005.07496]:

| Class               | Temporal Representation     | Canonical Examples                |
|---------------------|----------------------------|------------------------------------|
| Pseudo-dynamic      | Static weighted GNN (+decay) | TDGNN [Qu et al.]                 |
| Discrete-time       | Snapshot sequence           | GCRN-M1/M2, DySAT, EvolveGCN      |
| Continuous-time     | Event sequence              | JODIE, DyRep, TGAT, TGN           |

In discrete-time models, DGNNs are constructed by stacking a spatial GNN (e.g., GCN, GAT) with a temporal encoder (RNN, self-attention, etc.), or by integrating recurrent operations within the message passing itself.

## 2. Core Model Architectures

- **Stacked DGNNs:** Apply a spatial GNN at each time step to generate node-level embeddings $z_{i}^t$, feeding the output into a temporal model such as an LSTM or self-attention encoder:
  $$
  h^{t}_i = \mathrm{LSTM}(h^{t-1}_i, z^{t}_i)
  $$
  Example: GCRN-M1, DySAT [2005.07496].

- **Integrated DGNNs:** Fuse spatial and temporal updates, e.g., by replacing convolution with graph convolution in a convLSTM:
  $$
  f_t = \sigma(W_f *_\mathcal{G} X_t + U_f *_\mathcal{G} h_{t-1} + w_f \odot c_{t-1} + b_f)
  $$
  Example: GCRN-M2, EvolveGCN [2005.07496].

- **Continuous-time DGNNs:** Process individual interaction events and maintain per-node embeddings updated recursively, often including time-encoding for elapsed intervals:
  $$
  h_v^t = \mathrm{DGNN}(h_v^{t-}, \mathrm{AGG}(\{\text{neighbors at }<t\}))
  $$
  Example: JODIE, DyRep, TGN, TGAT [2005.07496], [2311.12255].

- **Transformer-based DGNNs:** Employ self-attention mechanisms (SAMs) to model sequential interactions, with advances regarding interaction-level chronology, mixed temporal encoding, and bidirectional interaction features. TIDFormer introduces a particularly interpretable SAM at the interaction level [2506.00431].

- **High-order and Overlap-aware Models:** Incorporate second-order and high-order structural signals such as neighborhood overlap directly in the message passing, as in NO-HGNN, to enrich representation power for tasks like link prediction [2506.06728].

- **Decoupled DGNNs:** Separate graph propagation and downstream sequence modeling, allowing one to use arbitrary predictors (e.g. LSTM, Transformer) on precomputed temporal node representations, increasing scalability [2305.08273].

## 3. Temporal and Structural Encoding Mechanisms

DGNNs incorporate temporal information through various mechanisms:

- **Explicit time encodings:** Functions (e.g., sinusoidal, time2vec, calendar-based) map elapsed intervals into feature vectors [2311.12255], [2506.00431].
- **Memory modules:** GRU, LSTM, or proprietary memory stacks maintain per-node temporal state, with architectures dictating whether spatial and temporal updates are independent, sequential, or integrated [1810.10627], [2311.12255].
- **Attention over history:** Self-attention layers (e.g., DySAT, TIDFormer) model the influence of past states and interactions, with proposals for interaction-level tokens leading to interpretable and robust models [2506.00431].
- **Message passing with adaptive forgetting:** Propagate information to neighbors according to recency, strength, and tie decay, as in streaming DGNNs [1810.10627].

Structural information is encoded using standard GNN aggregation functions, with advanced models supporting high-order (e.g., $k$-hop, overlap-aware) structures:

- **Tensor-based high-order propagation:** Third-mode tensor products and learnable time mixing enable efficient modeling of multi-hop and time cross-correlation [2506.06728].
- **Spectral decomposition:** DMD-GNNs employ data-driven, low-rank spectral filtering along principal dynamic modes recovered from sequences of node features, aligned with dynamical systems theory [2410.05593].

## 4. Training Objectives, Losses, and Performance Results

DGNNs are trained primarily for dynamic link prediction, node classification, and sometimes temporal forecasting. Standard objectives include:

- **Binary cross-entropy loss** for link prediction, often with negative sampling.
- **Cross-entropy loss** for node classification.
- **Temporal point process log-likelihood** for event generation in TPP-based models.
- **Contrastive pre-training** for temporal and structural discrimination (CPDG) using triplet-margin or auxiliary losses [2307.02813].
- **Information bottleneck regularization** to compress representations to the minimal sufficient consensus state and enhance robustness (DGIB) [2402.06716].
- **Spectral or operator regularization:** Encouraging dynamics learned by DMD modes to approximate observed feature transitions [2410.05593].

Empirical evaluations consistently show that DGNNs outperform both static GNNs and traditional dynamic embedding baselines across a range of datasets, tasks, and domains [2005.07496], [1810.10627], [2311.12255], [2305.08273], [2307.02813], [2410.05593], [2506.06728].

## 5. Scalability and Hardware Acceleration

Scalability is a critical consideration for DGNN deployment:

- **Decoupling propagation and prediction** enables the use of arbitrary sequence models and supports billion-edge graphs on a single machine [2305.08273].
- **Incremental computation and caching:** Frameworks such as ReInc systematically reuse intermediate GNN aggregations, cache temporal feature sequences, and exploit independence between sliding training windows to achieve order-of-magnitude speedups and near-linear distributed scaling [2501.15348].
- **FPGA and hardware acceleration:** DGNN-Booster demonstrates multi-level pipelining (across time steps and nodes) and adaptable dataflow architectures for DGNNs on FPGAs, achieving up to $8.4\times$ speedup over GPU and three orders of magnitude higher energy efficiency for inference [2304.06831].
- **Dynamic kNN evaluation and approximate neighborhood search:** In point cloud applications, dynamic graph construction (e.g., via EdgeConv) constitutes over 90% of inference latency, motivating hardware and algorithmic optimization [2309.09142].

## 6. Explainability, Robustness, and Pre-training

Explainability and robustness have emerged as principal concerns in DGNN research:

- **Explainability frameworks:** The DGExplainer system applies layer-wise relevance propagation (LRP) through both spatial and temporal layers to ascribe prediction influence to specific nodes, timestamps, and inputs, overcoming the limitations of static explainers [2207.11175].
- **Robustness under adversarial perturbations:** Information-theoretic DGIB models employ minimal-sufficient-consensual (MSC) conditions to compress and regularize temporal embeddings, providing resistance against targeted and untargeted attacks [2402.06716].
- **Self-supervised and contrastive pre-training:** Pre-training with temporal and structural contrast, as in CPDG, substantially enhances generalization and transfer across tasks, time splits, and even domains [2307.02813].

## 7. Current Challenges and Future Directions

Outstanding challenges for DGNNs include:

- **Scalability** to truly massive, rapidly evolving graphs, supporting both continuous and discrete streams.
- **Temporal granularity**: Adapting models to variable bin sizes and capturing relevant timescales, with coarse bins sometimes improving robustness by reducing timestamp noise [2311.12255].
- **Heterogeneous and high-order structures:** Generalizing to multi-relational, signed, or multiplex dynamic graphs, and incorporating high-order dependencies [2506.06728].
- **Model-agnostic uncertainty quantification**: Wrapper frameworks such as GSNOP provide ODE-based, uncertainty-aware prediction modules for DGNNs, particularly in sparse-data regimes [2211.08568].
- **Automated architecture and hyperparameter selection:** Efficient grid and policy search as complexity and dataset diversity grow.
- **Hardware-aware and out-of-core algorithms:** Integration of dataflow and cache optimizations for training and inference on resource-constrained devices [2501.15348], [2304.06831].

DGNN research continues to advance along axes of dynamic representation power, scalability, interpretability, robustness, and efficiency, presenting a rich landscape of architectures and methodologies for time-variant graph learning.

Source: https://www.emergentmind.com/topics/dynamic-graph-neural-networks-dgnns