Research map / Multi-agent systems and games / Cooperative multi-agent RL
RiskQ: Risk-sensitive Multi-Agent Reinforcement Learning Value Factorization
Siqi Shen, Chennan Ma, Chao Li, Weiquan Liu, Yongquan Fu, Songzhu Mei, Xinwang Liu, Cheng Wang
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 |
|---|---|---|---|
| make about 50% of agents use a different risk measure (w/o RIGM VaR_1, CVaR_1, CVaR_0.2) (ablation) | RiskQ · SMAC and MACN | test win rate / return | Hurt |
| replace attention mixer with sum of per-agent percentiles (RiskQ-Sum) (ablation) | RiskQ · SMAC and MACN | test win rate / return | Hurt |
| represent percentiles as expectation of state-action function (RiskQ-Qi) (ablation) | RiskQ · SMAC and MACN | test win rate / return | Hurt |
| use QMIX/residual mixers to remove representation limitation (RiskQ-QMIX, -Residual, -Residual-QMIX) (ablation) | RiskQ · 3s_vs_5z and MMM2 | test win rate | No clear effect |
| vary risk metric, risk level and number of percentiles (ablation) | RiskQ · SMAC | test win rate | Mixed |
| combine RiskQ with modified IQN Bellman operator for contraction (ablation) | RiskQ · SMAC | test win rate | Hurt |
| combine RiskQ with LQN risk-sensitive exploration (ablation) | RiskQ · SMAC | test win rate | No clear effect |
| use RiskQ instead of risk-sensitive baselines RMIX and DRIMA | RiskQ · MACN 4x11 and 4x15 | test return | Improved |
Closest papers by meaning
- Provable Risk-Sensitive Distributional Reinforcement Learning with General Function Approximation (ICML 2024 · similarity 0.89)
- Pessimism Meets Risk: Risk-Sensitive Offline Reinforcement Learning (ICML 2024 · similarity 0.88)
- Risk-Conditioned Reinforcement Learning: A Generalized Approach for Adapting to Varying Risk Measures (AAAI 2024 · similarity 0.88)
- Beyond CVaR: Leveraging Static Spectral Risk Measures for Enhanced Decision-Making in Distributional Reinforcement Learning (ICML 2025 · similarity 0.87)
- Breaking the Curse of Multiagency in Robust Multi-Agent Reinforcement Learning (ICML 2025 · similarity 0.87)
- Taming the Curses of Multiagency in Robust Markov Games with Large State Space through Linear Function Approximation (NeurIPS 2026 · similarity 0.87)
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.
