Reproduce “Dual-Primal Graph VAEs for Noisy Label Aggregation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Patrick Stinson, Nikolaus Kriegeskorte
The authors introduce DPGVAE, a graph-attention autoencoder that infers unknown crowd labels while reconstructing workers’ votes. Its encoder uses relationships among annotations, and its decoder predicts votes; inferred task-label distributions serve as answers without ground-truth training labels. Across eight benchmarks, it has the best average rank (2.1, versus 3.3), but not on every dataset; it lacks a formal identifiability guarantee.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.