Reproduce “Opening the Black Box of Neural Auction Mechanisms” 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
Li Kong, Zilong Liu, Haoyang Zhou, Hanbing Liu, Qi Qi
The authors audit a fixed CITransNet auction network by decomposing hidden states into sparse features and testing descriptions and interventions. In the primary two-bidder, five-item setting, 165 of 512 features met the authors’ criteria. In a separate two-bidder, single-item test, F62 identified Myerson’s no-sale region with AUC 0.9636, while interventions shifted oracle errors; this supports feature-based auditing, not exact implementation or generalization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.