AI Paper Atlas

Research map / Multi-agent systems and games / Games and equilibria

Deep Contract Design via Discontinuous Networks

Tonghan Wang, Paul Duetting, Dmitry Ivanov, Inbal Talgam-Cohen, David C. Parkes

NeurIPS 2023 · Multi-agent systems and games › Games and equilibria · approach: Learning with strategic agents · 2 citations (OpenAlex)

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 changedWhereMeasuredEffect
use DeLU instead of continuous ReLU networkcontract design network · Synthetic contract designoptimality: +28.29%Improved
use DeLU instead of picking best training contractcontract design network · Synthetic contract designoptimality: +23.84%Improved
DeLU on large problems with more than 1,000 outcomesDeLU · Large-scale contract designoptimality vs training data: at least 1.7 times betterImproved
use gradient-based inference instead of LP inference (ablation)DeLU · Synthetic contract designoptimality: -7.56%Hurt
use gradient-based inference instead of LP inference (ablation)DeLU · Inference time, m=25inference time: 50-90% less overheadImproved
initialize gradient inference with random contracts instead of training set (ablation)DeLU gradient inference · (5,16) problemoptimality: 95.29 -> 95.33No clear effect
initialize gradient inference with random contracts instead of training set (ablation)DeLU gradient inference · (32,50) problemoptimality: 88.43 -> 88.46No clear effect
increase independent cost weight beta_p from 0.0 to 0.9 at alpha_p=0.5DeLU · Cost correlation studyoptimality: 88.79 -> 97.23Improved
increase alpha_p from 0.5 to 0.9 at beta_p=0.6DeLU · Cost correlation studyoptimality: 95.22 -> 91.97Hurt
add sub-argmax inference to gradient-based inference (ablation)DeLU · Boundary alignment studyoptimalityMixed
higher boundary alignment degree between DeLU and true utilityDeLU LP inference · Boundary alignment studyoptimalityImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy