← Home

Daily arXiv Papers

Graph Neural Networks · Graph Learning · LLM × Graph
Conferences
Archives
2026-08-11 2026-08-10 2026-08-09 2026-08-08 2026-08-07 2026-08-06 2026-08-05 2026-08-04 2026-08-03 2026-08-02 2026-08-01 2026-07-31 2026-07-30 2026-07-29 2026-07-28 2026-07-27 2026-07-26 2026-07-25 2026-07-24 2026-07-23 2026-07-22 2026-07-21 2026-07-20 2026-07-19 2026-07-18 2026-07-17 2026-07-16 2026-07-15 2026-07-14 2026-07-13 2026-07-12 2026-07-11 2026-07-10 2026-07-09 2026-07-08 2026-07-07 2026-07-06 2026-07-05 2026-07-04 2026-07-03 2026-07-02 2026-07-01 2026-06-30 2026-06-29 2026-06-28 2026-06-27 2026-06-26 2026-06-25 2026-06-24 2026-06-23 2026-06-22 2026-06-21 2026-06-20 2026-06-19 2026-06-18 2026-06-17 2026-06-16 2026-06-15 2026-06-14 2026-06-13 2026-06-12 2026-06-11 2026-06-10 2026-06-09 2026-06-08 2026-06-07 2026-06-06 2026-06-05 2026-06-04 2026-06-03 2026-06-02 2026-06-01 2026-05-31 2026-05-30 2026-05-29 2026-05-28 2026-05-27 2026-05-26 2026-05-25 2026-05-24 2026-05-23 2026-05-22 2026-05-21 2026-05-20 2026-05-19 2026-05-18 2026-05-17 2026-05-16 2026-05-15 2026-05-14 2026-05-13 2026-05-12 2026-05-11 2026-05-10 2026-05-09 2026-05-08 2026-05-07 2026-05-06 2026-05-05 2026-05-04 2026-05-03 2026-05-02 2026-05-01 2026-04-30 2026-04-29 2026-04-28 2026-04-27 2026-04-26 2026-04-25 2026-04-24 2026-04-23 2026-04-22 2026-04-21 2026-04-20 2026-04-19 2026-04-18 2026-04-17 2026-04-16 2026-04-15 2026-04-14 2026-04-13 2026-04-12 2026-04-11 2026-04-10 2026-04-09 2026-04-08 2026-04-07 2026-04-06 2026-04-05 2026-04-04 2026-04-03 2026-04-02 2026-04-01 2026-03-31 2026-03-30 2026-03-29 2026-03-28 2026-03-27 2026-03-26 2026-03-25 2026-03-24 2026-03-23 2026-03-22 2026-03-21 2026-03-20 2026-03-19 2026-03-18 2026-03-17 2026-03-16 2026-03-15 2026-03-14 2026-03-13 2026-03-12 2026-03-11 2026-03-10 2026-03-09 2026-03-07 2026-03-06 2026-03-05 2026-03-04 2026-03-03 2026-03-02 2026-03-01 2026-02-28 2026-02-27 2026-02-26 2026-02-25 2026-02-24 2026-02-23 2026-02-22 2026-02-21 2026-02-20 2026-02-19 2026-02-18 2026-02-17 2026-02-16 2026-02-15 2026-02-14 2026-02-13 2026-02-12

Showing 27 papers for 2026-07-31

EvoCause: LLM-Guided Evolution of Causal Graphs for Root Cause Analysis
Graph Learning

EvoCause presents an LLM-guided framework to evolve causal graphs for root cause analysis in alarm cascades. It uses expert labels to constrain source alarms and dynamically updates the graph across historical incidents, improving RCA accuracy.

THGFM: Dual-Branch Temporal Heterogeneous Graph Fusion Model
GNN Graph Learning

THGFM introduces a dual-branch model for temporal heterogeneous graphs that jointly captures cross-type heterogeneity and temporal dynamics. It integrates time within the attention kernel to enable parameter-efficient transfer and scalable learning on web-scale data.

Schreier-Coset Graph Rewiring
GNN Graph Learning

The paper proposes Schreier-Coset Graph Rewiring, a group-theoretic graph rewiring method that alleviates over-squashing while preserving key graph properties and avoiding edge explosion.

FedOGL: Combating Catastrophic Forgetting in Federated Open-World Multimodal Graph Learning
Graph Learning

FedOGL tackles catastrophic forgetting in federated open-world multimodal graph learning. It enables clients to learn emerging classes from private multimodal streams while preserving past knowledge and rejecting out-of-distribution samples via federated updates.

Train Small, Deploy Large: Zero-Shot GNN Transfer Through Geometric Renormalization
GNN Graph Learning

We propose a zero-shot transfer protocol where a GNN trained on a geometrically renormalized (coarse) graph can be deployed on the full-resolution graph without retraining. We show conditions under which transferred weights work and demonstrate performance gains.

What Makes Graph Unified? Principles and Generative Sliding-Window Transformer for Graph Foundation Models
Graph Learning

We discuss principles for Graph Foundation Models and present a generative sliding-window Transformer for graphs that can handle heterogeneous features. The model processes graphs as sequences to capture multi-scale patterns.

Persistent Gaussian Perturbations Prevent Oversmoothing in Recurrent Graph Neural Networks
GNN Graph Learning

We study a recurrent GNN with independent Gaussian noise injected after every propagation step to combat oversmoothing. The noise acts as persistent perturbations and preserves representation diversity, with theoretical and empirical support.

TopoFormer: Topology Meets Attention for Graph Learning
Graph Learning

TopoFormer encodes topology into attention-friendly sequences using Topo-Scan, creating topological tokens via node/edge filtrations. A Transformer processes these sequences to produce graph-level embeddings that capture local to global structure.

Semi-Supervised Learning for Molecular Graphs via Ensemble Consensus
GNN Graph Learning

The authors propose an ensemble-consensus semi-supervised method for molecular property prediction that leverages unlabeled data by achieving robust agreement across diverse augmentations/models. This improves performance under limited labels.

Same Graph Cross-Task Transfer in GNNs: Protocols and Predictors
GNN Graph Learning

The paper formalizes same-graph cross-task transfer between node classification and link prediction and introduces a leakage-free protocol and predictors to reliably evaluate transfer on the same underlying graph.

Dynamic Spectral Filtering for Temporal Graph Learning: Learning Evolving Propagation Operators
Graph Learning

Dynamic Spectral Filtering (DSF) models evolving propagation operators for temporal graphs by using time-dependent Chebyshev filters with vector-valued coefficients treated as recurrent states. A recurrent branch updates these coefficients over time.

Fully Inductive Cardinality Estimation
Knowledge Graph

FICE introduces the first fully inductive learned estimator for cardinalities of Basic Graph Pattern queries over knowledge graphs, generalizing to unseen graphs without retraining.

Graph Neural Multilevel Preconditioners for Iterative Solvers
GNN Graph Learning

This work develops Graph Neural Multilevel Preconditioners to enhance iterative solvers by combining AMG-style hierarchies with GNN-based improvements, improving robustness on challenging systems.

Graph Neural Network Force Fields for Spin Dynamics in Metallic Magnets
GNN Graph Learning

A GNN-based magnetic force-field learns the effective energy functional governing itinerant spin dynamics from electronic calculations, enabling fast simulations of spin evolution in metallic magnets.

ScaFE: Data-Efficient Scar Classification with LLM-Generated Clinical Feature Programs
Graph Learning

ScaFE transfers clinical knowledge from a large language model to generate clinical feature programs for scar classification, enabling data-efficient classification from photographs. It reduces reliance on large labeled datasets.

Dissecting Federated-Graph Aggregation under Domain Shift: Importance-Aware Aggregation via Empirical Analysis
Graph Learning

The paper analyzes how graph-domain shifts impact federated graph aggregation and shows that simple one-size-fits-all aggregation is suboptimal. It advocates importance-aware, coordinate-wise aggregation to improve robustness under domain shifts.

ARES: Anomaly Recognition Model For Edge Streams
Graph Learning

ARES proposes an anomaly recognition model for edge streams in temporal graphs, enabling real-time detection of unusual edge events and addressing concept drift and high data volumes with scalable methods.

MSGNN: A Spectral Graph Neural Network Based on a Novel Magnetic Signed Laplacian
GNN Graph Learning

MSGNN introduces a spectral GNN based on the magnetic signed Laplacian, generalizing the signed and magnetic Laplacians for signed-directed graphs, and builds a corresponding efficient GNN architecture.

A Methodology for Designing Knowledge-Driven Missions for Robots
Knowledge Graph

The paper presents a methodology for implementing knowledge graphs in ROS 2 systems to enhance autonomous robotic missions, detailing steps from initial and target conditions to task planning and high-level mission design.

Automorphism-Induced Non-Canonicity in Top-k Explanations of Graph Neural Networks
GNN Graph Theory

This work analyzes automorphism-induced non-canonicity in top-k explanations of gradient-based GNNs, showing that graph symmetries can make attribution ties and top-k selection depend on input ordering; it discusses structural obstructions and suggests canonicalization-based fixes.

AgentGFM: A Graph Foundation Model with Node-Agent Information-Flow Control
Graph Learning

AgentGFM proposes a Graph Foundation Model that uses node-agent information-flow control to adapt message passing to local graph structure, addressing heterogeneity across graphs and within graphs. Unlike fixed propagation schemes, it learns to tailor information flow, improving the transferability of topological patterns to unseen graphs.

Guarding Organizations Against Malware Risk: A Novel Graph-Based Malware Detection Method
Graph Learning

A novel graph-based malware detection method is proposed to counter evasion by byte-level manipulation. By representing software as program graphs that capture execution behavior, the method achieves robustness beyond traditional byte-based ML, improving accuracy and resilience to evasive techniques.

TRACE-CTI: Auditable Post-Extraction Governance of TTP Claims with Knowledge Graphs
Knowledge Graph

TRACE-CTI presents an auditable post-extraction governance framework for TTP claims using knowledge graphs. It preserves run-level predictions, aggregates them into configuration-level GraphAssertions, and materializes corroboration as ConsensusAssertions, enabling auditable mappings with provenance and validation history for MITRE ATT&CK.

Detecting Knowledge Inconsistencies Across Text, Tables, and Knowledge Graphs
Knowledge Graph

We study cross-modal inconsistencies among text, tables, and knowledge graphs in Wikipedia and Wikidata. We propose a taxonomy of cross-modal knowledge inconsistencies and a detection method that handles granularity differences, direct conflicts, and temporal changes, along with explanations of conflicts.

An Exploration Graph with Continuous Refinement for Efficient Multimedia Retrieval
Graph Learning

An Exploration Graph with Continuous Refinement (crE) for efficient multimedia retrieval is proposed. It uses a graph-based ANNS framework with a continuously refining exploration graph to address growing datasets, achieving fast index construction and lower memory while preserving retrieval accuracy.

Dynamic Exploration Graph: A Novel Approach for Efficient Nearest Neighbor Search in Evolving Multimedia Datasets
Graph Learning

Dynamic Exploration Graph (DEG) extends the continuous refining Exploration Graph to dynamic datasets. It retains high search efficiency for static data while adding support for data additions and removals, enabling efficient updates and maintaining accuracy.

GLM-RAG: Graph Language Models for Graph-Based Retrieval-Augmented Generation
GNN Graph Learning Knowledge Graph LLM × Graph

GLM-RAG introduces a Graph Language Model-based retriever for graph-based Retrieval-Augmented Generation. It investigates the strengths of GLM-based retrievers compared with GNN-based methods and demonstrates improved retrieval and reasoning capabilities on graph-augmented generation tasks.