Research map / Multi-agent systems and games / Games and equilibria
Deep Incentive Design with Differentiable Equilibrium Blocks
Vinzenz Thoma, Georgios Piliouras, Luke Marris
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 |
|---|---|---|---|
| evaluate with exact ECOS equilibrium instead of the DEB (ablation) | DID · Contract design (CE) | change in principal utility: 0.25 -> 0.05 | Hurt |
| evaluate with exact ECOS equilibrium instead of the DEB (ablation) | DID · Contract design (CCE) | change in principal utility: 0.33 -> 0.20 | Hurt |
| evaluate with exact ECOS equilibrium instead of the DEB (ablation) | DID · Inverse equilibrium (CE) | KL to target joint: 0.056 -> 0.175 | Hurt |
| evaluate with exact ECOS equilibrium instead of the DEB (ablation) | DID · Inverse equilibrium (CCE) | KL to target joint: 0.043 -> 0.071 | Hurt |
| evaluate with exact ECOS equilibrium instead of the DEB (ablation) | DID · Machine scheduling (CE) | change in makespan: -0.339 -> -0.096 | Hurt |
| evaluate with exact ECOS equilibrium instead of the DEB (ablation) | DID · Machine scheduling (CCE) | change in makespan: -0.211 -> -0.050 | Hurt |
| add local Nelder-Mead polishing to DID (ablation) | DID · Contract design (CE, DEB) | change in principal utility: 0.25 -> 0.51 | Improved |
| add local polishing to DID (ablation) | DID · Contract design (CCE, DEB) | change in principal utility: 0.33 -> 0.59 | Improved |
| add local polishing to DID (ablation) | DID · Inverse equilibrium (CCE, DEB) | KL to target joint: 0.043 -> 0.013 | Improved |
| add local polishing to DID (ablation) | DID · Inverse equilibrium (CE, DEB) | KL to target joint: 0.056 -> 0.024 | Improved |
| add local polishing to DID (ablation) | DID · Machine scheduling (CCE, DEB) | change in makespan: -0.211 -> -0.240 | Improved |
| add local polishing to DID (ablation) | DID · Machine scheduling (CE, ECOS) | change in makespan: -0.096 -> -0.352 | Improved |
Closest papers by meaning
- Scalable Neural Incentive Design with Parameterized Mean-Field Approximation (NeurIPS 2025 · similarity 0.90)
- Multi-Sender Persuasion: A Computational Perspective (ICML 2024 · similarity 0.85)
- Explicit Exploration for High-Welfare Equilibria in Game-Theoretic Multiagent Reinforcement Learning (ICML 2025 · similarity 0.85)
- Learning Equilibria in Adversarial Team Markov Games: A Nonconvex-Hidden-Concave Min-Max Optimization Problem (NeurIPS 2024 · similarity 0.85)
- Aligning Individual and Collective Objectives in Multi-Agent Cooperation (NeurIPS 2024 · similarity 0.84)
- Deviate or Not: Learning Coalition Structures with Multiple-bit Observations in Games (AAAI 2025 · similarity 0.84)
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.
