← all datasets

BeGIN

Emerging
1papers using it
2025first seen

BeGIN (Benchmarking for Graphs with Instance-dependent Noise) is a benchmark that provides realistic graph datasets with various noise types to evaluate noise-handling strategies across Graph Neural Network architectures, noisy label detection, and noise-robust learning.

🔎 Find this dataset

Papers using BeGIN (1)

BeGIN — datasets — graph-learning