SAGE is the first end-to-end LLM-driven multi-agent fraud detection system using a Data Diagnostic Tree and MDP optimization, achieving 40.86% average F1 gain and winning 96% of comparisons across five datasets and five backbones.
Fraud dataset benchmark and applications.arXiv preprint arXiv:2208.14417, 2022
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
TFMs achieve top AUC on TALENT but lower conditional coverage (SSCS) than GBDTs under conformal prediction, revealing a performance-uncertainty trade-off.
SilIF blends Isolation Forest path lengths with silhouette scores from clustered fingerprints, yielding +0.008 AUC-PR gain on IEEE-CIS fraud data but no gain on synthetic credit-card data.
citing papers explorer
-
SAGE: An LLM-driven Self Reflective Agentic Framework for Fraud Detection
SAGE is the first end-to-end LLM-driven multi-agent fraud detection system using a Data Diagnostic Tree and MDP optimization, achieving 40.86% average F1 gain and winning 96% of comparisons across five datasets and five backbones.
-
High Performance, Low Reliability: Uncertainty Benchmarking for Tabular Foundation Models
TFMs achieve top AUC on TALENT but lower conditional coverage (SSCS) than GBDTs under conformal prediction, revealing a performance-uncertainty trade-off.
-
SilIF: Silhouette-Augmented Isolation Forest for Unsupervised Transaction Fraud Detection
SilIF blends Isolation Forest path lengths with silhouette scores from clustered fingerprints, yielding +0.008 AUC-PR gain on IEEE-CIS fraud data but no gain on synthetic credit-card data.