# Amortised Post-Hoc Explanation with Exact Preservation for Dynamic Graph Anomaly Detectors

arXiv: https://arxiv.org/abs/2608.15559 · HTML: https://mlsanity.com/paper/2608.15559/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 19 references that are in the map.
- First posted on arXiv 16 Aug 2026.

## Builds on (most cited first, 12 of 12)

- [Semi-Supervised Classification with Graph Convolutional Networks](https://mlsanity.com/paper/1609.02907/) (2016)
- [Categorical Reparameterization with Gumbel-Softmax](https://mlsanity.com/paper/1611.01144/) (2016)
- [Sanity Checks for Saliency Maps](https://mlsanity.com/paper/1810.03292/) (2018)
- [The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables](https://mlsanity.com/paper/1611.00712/) (2016)
- [Link Prediction Based on Graph Neural Networks](https://mlsanity.com/paper/1802.09691/) (2018)
- [Parameterized Explainer for Graph Neural Network](https://mlsanity.com/paper/2011.04573/) (2020)
- [Explainability in Graph Neural Networks: A Taxonomic Survey](https://mlsanity.com/paper/2012.15445/) (2020)
- [Anomaly Detection in Dynamic Graphs via Transformer](https://mlsanity.com/paper/2106.09876/) (2021)
- [Structural Temporal Graph Neural Networks for Anomaly Detection in Dynamic Graphs](https://mlsanity.com/paper/2005.07427/) (2020)
- [TempME: Towards the Explainability of Temporal Graph Neural Networks via Motif Discovery](https://mlsanity.com/paper/2310.19324/) (2023)
- [Anomaly Detection in Dynamic Graphs: A Comprehensive Survey](https://mlsanity.com/paper/2406.00134/) (2024)
- [SLADE: Detecting Dynamic Anomalies in Edge Streams without Labels via Self-Supervised Learning](https://mlsanity.com/paper/2402.11933/) (2024)

## Abstract (arXiv)

Anomaly detection in dynamic graphs underpins financial fraud analysis, intrusion detection, and platform integrity, where automated decisions require human-interpretable justifications. StrGNN, the strongest performer in recent benchmarks, produces no explanation: when an edge is flagged, the analyst receives only a score. Explanation metrics are undefined for StrGNN because no attribution vector exists. This paper closes that gap. We present X-StrGNN, a post-hoc explanation layer that wraps a trained, frozen StrGNN and emits, for every flagged edge, dual attributions: a structural attribution identifying which contextual interactions in the enclosing subgraph drove the decision, and a temporal attribution identifying which historical snapshot carried the signal. Both attributions are multiplicative masks identically one in the unexplained pass, so the layer is an exact pass-through: detection is preserved to machine precision, verified rather than asserted (Delta AUC = 0.0000, Delta AP = 0.0000, Delta P@100 = 0.0000). Attribution costs 0.66 ms per edge, making explanation of an entire alarm list feasible. We conduct the first controlled design study of attribution strategies for this architecture, comparing gradient attribution, per-instance mask optimisation, and amortised parameterisation under one protocol, one budget, and three seeds. X-StrGNN attains the highest stability (0.913) at 268x lower cost than per-instance optimisation, and its temporal attribution (1.601 against a measured random floor of 0.973) is separably better than its ablated control, while per-instance optimisation - the most expensive strategy - falls below that floor. Code, protocol, and per-seed measurements are released.