# A Scalable Neural Network for DSIC Affine Maximizer Auction Design Source: [A Scalable Neural Network for DSIC Affine Maximizer Auction Design]() Collection index: [llms.txt]() Data updated: 2026-01-02T04:46:16.657Z Authors: Zhijian Duan, Haoran Sun, Yurong Chen, Xiaotie Deng Published: 2023-05-20T10:42:00.000Z ## Abstract Automated auction design aims to find empirically high-revenue mechanisms through machine learning. Existing works on multi item auction scenarios can be roughly divided into RegretNet-like and affine maximizer auctions (AMAs) approaches. However, the former cannot strictly ensure dominant strategy incentive compatibility (DSIC), while the latter faces scalability issue due to the large number of allocation candidates. To address these limitations, we propose AMenuNet, a scalable neural network that constructs the AMA parameters (even including the allocation menu) from bidder and item representations. AMenuNet is always DSIC and individually rational (IR) due to the properties of AMAs, and it enhances scalability by generating candidate allocations through a neural network. Additionally, AMenuNet is permutation equivariant, and its number of parameters is independent of auction scale. We conduct extensive experiments to demonstrate that AMenuNet outperforms strong baselines in both contextual and non-contextual multi-item auctions, scales well to larger auctions, generalizes well to different settings, and identifies useful deterministic allocations. Overall, our proposed approach offers an effective solution to automated DSIC auction design, with improved scalability and strong revenue performance in various settings. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Contextual Auction on Setting (A) 3x2](): Contextual Auction; Setting (A) 3x2; Average Revenue: 56.01. - [Contextual Auction on Setting (A) 2x2 1.0 (test)](): Contextual Auction; Setting (A) 2x2 1.0 (test); Average Revenue: 0.4398. - [Contextual Auction on Setting (A) 2x5](): Contextual Auction; Setting (A) 2x5; Average Revenue: 1.1539. - [Contextual Auction on Setting (A) 2x10](): Contextual Auction; Setting (A) 2x10; Average Revenue: 2.3935. - [Contextual Auction on Setting (A) 3x5](): Contextual Auction; Setting (A) 3x5; Average Revenue: 1.4458. - [Contextual Auction on Setting (A) 3x10](): Contextual Auction; Setting (A) 3x10; Average Revenue: 2.9006. - [Contextual Auction on Setting (B) 4x2](): Contextual Auction; Setting (B) 4x2; Average Revenue: 70.09. - [Contextual Auction on Setting B 5x2](): Contextual Auction; Setting B 5x2; Average Revenue: 0.7548. - [Contextual Auction on Setting (B) 6x2](): Contextual Auction; Setting (B) 6x2; Average Revenue: 0.8127. - [Contextual Auction on Setting (B) 7x2](): Contextual Auction; Setting (B) 7x2; Average Revenue: 0.856. - [Auction Revenue Maximization on Classic auction Setting (F) 1x2](): Auction Revenue Maximization; Classic auction Setting (F) 1x2; Average Revenue: 17.01. - [Auction Revenue Maximization on Classic auction Setting (C) 5x5](): Auction Revenue Maximization; Classic auction Setting (C) 5x5; Avg Revenue: 3.3916. - [Auction Revenue Maximization on Classic auction Setting (D) 3x1](): Auction Revenue Maximization; Classic auction Setting (D) 3x1; Average Revenue: 2.7382. - [Auction Revenue Maximization on Classic auction Setting (E) 1x2](): Auction Revenue Maximization; Classic auction Setting (E) 1x2; Avg Revenue: 9.6219. - [Optimal Auction Design on 3x10 auction setting](): Optimal Auction Design; 3x10 auction setting; Revenue: 5.5896. - [Optimal Auction Design on 2x5 auction setting](): Optimal Auction Design; 2x5 auction setting; Revenue: 2.2768. ## Original sources and implementation [Original paper]() [PDF]() [Implementation]()