Attention-based graph neural networks: a survey

2026-05-12T08:52:13Z4ba95ce959571b5f3103915eca520de3702dbb979e81c648881cae4d5f8e7e4b
LLMadversarial-aiattentionattributionbenchmarkbioinformaticscausal-inferencecreator-economydigital-twinfront-doorgraph-generationgraph-neural-networksgraph-transformermodel-safetyneuron-editingnode-insertionrecommender-systemsself-debiasingsurveysynthetic-networksvisualization

What happened

This RSS bundle contains multiple new arXiv publications (May 12, 2026) across graph ML, causal attribution in recommender systems, AI-human opinion dynamics, model safety, visualization, and trust in digital twin systems. Highlights: an up-to-date survey of attention-based GNNs (including graph transformers); ALM-MTA, a front-door causal multi-touch attribution approach applied at massive scale for creator-economy optimization; Astro Generative Network for controlled node insertion into observed graphs; GraphInstruct, a progressive benchmark diagnosing LLM graph-generation failures; MiRA, aブラ

Why it matters

A reviewed impact interpretation has not been published for this record.

Evidence and limitations

Source ID
arxiv_cs_si
Record identifier
4ba95ce959571b5f3103915eca520de3702dbb979e81c648881cae4d5f8e7e4b
Enrichment time
2026-05-12T08:52:13Z
AI-assisted enrichment
Yes

This record may overlap with other records. Its enrichment can be incomplete or wrong, and machine assistance was used. Validate consequential decisions against the linked source and your own environment.