Research map / Multi-agent systems and games / Cooperative multi-agent RL
N$\text{A}^\text{2}$Q: Neural Attention Additive Model for Interpretable Multi-Agent Q-Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use unary and pairwise interaction shape functions instead of only order-1 interactions (ablation) | NA²Q · SMAC hard and super hard scenarios | performance | Helped |
| Increase the interaction order beyond pairwise (ablation) | NA²Q · SMAC hard and super hard scenarios | interpretability | Hurt |
| Replace identity semantics with the global state (ablation) | NA²Q · SMAC hard and super hard scenarios | performance | Hurt |
| Remove attention from credit assignment (ablation) | NA²Q · SMAC hard and super hard scenarios | performance | Hurt |
| Remove the intervention function's global-state information (ablation) | NA²Q · SMAC hard and super hard scenarios | performance | Hurt |
| Add higher-order interactions beyond order 1 (ablation) | NA²Q-w/o-semantic · SMAC hard and super hard scenarios | performance | Helped |
| Use NA²Q instead of Qatten | LBF · LBF constructed tasks | test return | Helped |
| Use NA²Q instead of WQMIX or SHAQ | LBF · LBF constructed tasks | test return | No clear effect |
| Use NA²Q instead of VDN or QMIX | SMAC · Super hard scenarios | test win rate | Helped |
Closest papers by meaning
- High-order Interactions Modeling for Interpretable Multi-Agent Q-Learning (NeurIPS 2025 · similarity 0.89)
- Towards Understanding Cooperative Multi-Agent Q-Learning with Value Factorization (NeurIPS 2021 · similarity 0.88)
- Attention-Guided Contrastive Role Representations for Multi-agent Reinforcement Learning (ICLR 2024 · similarity 0.86)
- Enhancing Cooperative Multi-Agent Reinforcement Learning with State Modelling and Adversarial Exploration (ICML 2025 · similarity 0.86)
- Towards Efficient Collaboration via Graph Modeling in Reinforcement Learning (AAAI 2025 · similarity 0.86)
- Inverting the Bellman Equation: From $Q$-Values to World Models (NeurIPS 2026 · 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.