Research map / Learning theory and algorithms / Submodular and game algorithms
Certifying Strategyproof Auction Networks
Michael Curry, Ping-yeh Chiang, Tom Goldstein, John Dickerson
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add ReLU stability regularization (IR architecture) (ablation) | IR network · Auction 1x2 | solve time (s): 25.6 -> 7.2 | Improved |
| add ReLU stability regularization (IR architecture) (ablation) | IR network · Auction 2x2 | solve time (s): 13.9 -> 5.8 | Improved |
| add ReLU stability regularization (IR architecture) (ablation) | IR network · Auction 2x2 (2nd agent) | solve time (s): 17.4 -> 7.520 | Improved |
| add ReLU stability regularization (ablation) | IR network · Auction 1x2 | certified regret: 0.019 -> 0.004 | Improved |
| add ReLU stability regularization (ablation) | IR network · Auction 2x2 | certified regret: 0.014 -> 0.013 | Improved |
| add ReLU stability regularization (ablation) | IR network · Auction 1x2 | revenue: 0.593 -> 0.569 | Mixed |
| use non-IR architecture (distilled) instead of IR architecture (ablation) | non-IR network · Auction 1x2 | solve time (s): 7.2 -> 0.034 | Improved |
| use non-IR architecture (distilled) instead of IR architecture (ablation) | non-IR network · Auction 2x2 | solve time (s): 5.8 -> 5.480 | Improved |
| non-IR architecture instead of IR architecture (ablation) | non-IR network · Auction 1x2 | certified regret: 0.004 -> 0.011 | Hurt |
| non-IR architecture instead of IR architecture (ablation) | non-IR network · Auction 2x2 | certified regret: 0.013 -> 0.011 | Improved |
| enforce IR after the fact by filtering IR-violating points (ablation) | non-IR network · Auction 1x2 | revenue: 0.5738 -> 0.5681 | Hurt |
| enforce IR after the fact by filtering IR-violating points (ablation) | non-IR network · Auction 2x2 | revenue: 0.8874 -> 0.8824 | Hurt |
Closest papers by meaning
- Learning Optimal Auctions with Correlated Value Distributions (AAAI 2025 · similarity 0.89)
- Safely Learning Optimal Auctions: A Testable Learning Framework for Mechanism Design (ICML 2025 · similarity 0.88)
- On the Power of Randomization for Obviously Strategy-Proof Mechanisms (AAAI 2025 · similarity 0.88)
- Comparing Uniform Price and Discriminatory Multi-Unit Auctions through Regret Minimization (NeurIPS 2025 · similarity 0.87)
- Learning Safe Strategies for Value Maximizing Buyers in Uniform Price Auctions (ICML 2025 · similarity 0.86)
- AuctionNet: A Novel Benchmark for Decision-Making in Large-Scale Games (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
