mlsanityData to 14 Oct 2026

FedLSG: LLM-Enhanced Semantic Calibration for Federated Graph Backdoor Defense

arXiv 2607.19674 ↗First posted 22 Jul 2026cs.CR cs.AI cs.LGcited by 0

On the map: LLM safety and security → Backdoor and data poisoning attacks

Abstract (arXiv)

Federated Graph Neural Networks (FedGNNs) are highly vulnerable to backdoor poisoning, yet existing defenses typically rely on rule-based approaches that lack semantic understanding, making them vulnerable to stealthy triggers and harmful to benign structures. To solve this, we present FedLSG, the first framework that integrates large language models (LLMs) into federated graph backdoor defense. FedLSG introduces a graph and behavior to text grounding scheme that transforms local graph structures and client update behaviors into semantically rich natural language representations. The framework further adopts a lightweight student-teacher architecture. On the server side, a full scale LLM serves as a teacher, providing global contextual guidance and evaluating client updates during aggregation to identify potentially malicious participants. On the client side, a LoRA-based student is maintained to perform semantic reasoning, to suppress the influence of edges associated with backdoor triggers. By enabling semantic interpretation of both graph patterns and client behaviors, the framework adaptively incorporates rule-based signals into message passing and client aggregation for defense. Experiments demonstrate that FedLSG significantly improves resistance to backdoor attacks without compromising graph integrity.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security. Builds on 23 of its 39 references that are in the map. First posted on arXiv 22 Jul 2026.

Builds on

23 of 39 references.

References · 23

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2017Graph Attention Networkscited by 8,923
  3. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,163
  4. 2019GraphSAINT: Graph Sampling Based Inductive Learning Methodcited by 618
  5. 2020GNNGuard: Defending Graph Neural Networks against Adversarial Attackscited by 195
  6. 2020Backdoor Attacks to Graph Neural Networkscited by 128
  7. 2022Neurotoxin: Durable Backdoors in Federated Learningcited by 98
  8. 2020Defending against Backdoors in Federated Learning with Robust Learning Ratecited by 82
  9. 2023Unnoticeable Backdoor Attacks on Graph Neural Networkscited by 45
  10. 2022FederatedScope-GNN: Towards a Unified, Comprehensive and Efficient Package for Federated Graph Learningcited by 43
  11. 2023pFedLoRA: Model-Heterogeneous Personalized Federated Learning with LoRA Tuningcited by 40
  12. 2023Backdoor Attacks and Defenses in Federated Learning: Survey, Challenges and Future Research Directionscited by 33
11 more
  1. 2024Rethinking Graph Backdoor Attacks: A Distribution-Preserving Perspectivecited by 22
  2. 2022More is Better (Mostly): On the Backdoor Attacks in Federated Graph Neural Networkscited by 15
  3. 2022GARNET: Reduced-Rank Topology Learning for Robust and Scalable Graph Neural Networkscited by 14
  4. 2025FedALT: Federated Fine-Tuning through Adaptive Local Training with Rest-of-World LoRAcited by 13
  5. 2024A Simple and Yet Fairly Effective Defense for Graph Neural Networkscited by 12
  6. 2024Beyond Redundancy: Information-aware Unsupervised Multiplex Graph Structure Learningcited by 12
  7. 2024Personalized Federated Fine-Tuning for LLMs via Data-Driven Heterogeneous Model Architecturescited by 12
  8. 2024Robustness Inspired Graph Backdoor Defensecited by 7
  9. 2023Bkd-FedGNN: A Benchmark for Classification Backdoor Attacks on Federated Graph Neural Networkcited by 4
  10. 2025Data-centric Federated Graph Learning with Large Language Modelscited by 3
  11. 2025LLM-Guided Dynamic-UMAP for Personalized Federated Graph Learningcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026FedGraph: Defending Federated Large Language Model Fine-Tuning Against Backdoor Attacks via Graph-Based Aggregation
  2. 2024Distributed Backdoor Attacks on Federated Graph Learning and Certified Defenses
  3. 2026Beyond Corner Patches: Semantics-Aware Backdoor Attack in Federated Learning
  4. 2025Stealthy Dual-Trigger Backdoors: Attacking Prompt Tuning in LM-Empowered Graph Foundation Models
  5. 2026Revisiting Backdoor Threat in Federated Instruction Tuning from a Signal Aggregation Perspective
  6. 2026Rethinking Federated Graph Foundation Models: A Graph-Language Alignment-based Approach
  7. 2023G2uardFL: Safeguarding Federated Learning Against Backdoor Attacks through Attributed Client Graph Clustering
  8. 2024Backdoor Attack on Vertical Federated Graph Neural Network Learning