# Scalable and Efficient Joint Spiking Embedding Predictive Architecture for Large-Scale Dynamic Graphs

arXiv: https://arxiv.org/abs/2607.18412 · HTML: https://mlsanity.com/paper/2607.18412/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 47 references that are in the map.
- First posted on arXiv 20 Jul 2026.

## Builds on (most cited first, 15 of 15)

- [V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning](https://mlsanity.com/paper/2506.09985/) (2025)
- [On the Bottleneck of Graph Neural Networks and its Practical Implications](https://mlsanity.com/paper/2006.05205/) (2020)
- [Revisiting Feature Prediction for Learning Visual Representations from Video](https://mlsanity.com/paper/2404.08471/) (2024)
- [Inductive Representation Learning on Temporal Graphs](https://mlsanity.com/paper/2002.07962/) (2020)
- [Graph Neural Networks with Learnable Structural and Positional Representations](https://mlsanity.com/paper/2110.07875/) (2021)
- [Inductive Representation Learning in Temporal Networks via Causal Anonymous Walks](https://mlsanity.com/paper/2101.05974/) (2021)
- [VL-JEPA: Joint Embedding Predictive Architecture for Vision-language](https://mlsanity.com/paper/2512.10942/) (2025)
- [A-JEPA: Joint-Embedding Predictive Architecture Can Listen](https://mlsanity.com/paper/2311.15830/) (2023)
- [Spiking Graph Convolutional Networks](https://mlsanity.com/paper/2205.02767/) (2022)
- [Next-Embedding Prediction Makes Strong Vision Learners](https://mlsanity.com/paper/2512.16922/) (2025)
- [Exploiting Spiking Dynamics with Spatial-temporal Feature Normalization in Graph Learning](https://mlsanity.com/paper/2107.06865/) (2021)
- [Rethinking JEPA: Compute-Efficient Video SSL with Frozen Teachers](https://mlsanity.com/paper/2509.24317/) (2025)
- [Self-Supervised Dynamic Graph Representation Learning via Temporal Subgraph Contrast](https://mlsanity.com/paper/2112.08733/) (2021)
- [DyG2Vec: Efficient Representation Learning for Dynamic Graphs](https://mlsanity.com/paper/2210.16906/) (2022)
- [GT-SNT: A Linear-Time Transformer for Large-Scale Graphs via Spiking Node Tokenization](https://mlsanity.com/paper/2504.11840/) (2025)

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

- [Dynamic Graph World Modelling: Learning Future Dynamics from Temporal Context](https://mlsanity.com/paper/2610.19364/) (2026)

## Abstract (arXiv)

Dynamic graph learning aims to capture evolving structural and semantic patterns in real-world systems, such as fraud detection and recommender systems. Due to the scarcity of labeled data in real-world dynamic graphs, recent studies have introduced generative or contrastive paradigms (e.g., masked graph autoencoders or graph contrastive learning) to generate task-agnostic graph embeddings. However, these methods typically rely on complex edge-level reconstruction objectives and tailored graph augmentation strategies. This incurs substantial computational overhead when scaling to large-scale dynamic graphs. In this paper, we propose SG-JEPA, a joint spiking embedding predictive architecture for large-scale dynamic graphs. In contrast to existing self-supervised methods, SG-JEPA partitions nodes into context and target sets along the temporal dimension to learn embeddings that are predictive of each other via additional spatial-temporal information. Furthermore, through encoding sequential inputs into coarse-to-fine spike count embeddings, spiking neurons enable SG-JEPA to adapt to the varying computational constraints of downstream tasks. Extensive experiments demonstrate that SG-JEPA achieves competitive or even superior performance over discriminative baselines on node classification, while effectively scaling to the dynamic graph with 13 million edges. SG-JEPA avoids the complex machinery (negative sampling, graph augmentations, edge-level reconstruction, etc.), resulting in superior training efficiency and memory scalability compared with prior self-supervised dynamic graph baselines.