# Redistributing Rewards Across Time and Agents for Multi-Agent Reinforcement Learning

arXiv: https://arxiv.org/abs/2502.04864 · HTML: https://mlsanity.com/paper/2502.04864/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 28 references that are in the map.
- First posted on arXiv 7 Feb 2025.

## Builds on (most cited first, 23 of 23)

- [Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments](https://mlsanity.com/paper/1706.02275/) (2017)
- [Curiosity-driven Exploration by Self-supervised Prediction](https://mlsanity.com/paper/1705.05363/) (2017)
- [The Surprising Effectiveness of PPO in Cooperative, Multi-Agent Games](https://mlsanity.com/paper/2103.01955/) (2021)
- [Counterfactual Multi-Agent Policy Gradients](https://mlsanity.com/paper/1705.08926/) (2017)
- [The StarCraft Multi-Agent Challenge](https://mlsanity.com/paper/1902.04043/) (2019)
- [Value-Decomposition Networks For Cooperative Multi-Agent Learning](https://mlsanity.com/paper/1706.05296/) (2017)
- [Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2003.08839/) (2020)
- [Google Research Football: A Novel Reinforcement Learning Environment](https://mlsanity.com/paper/1907.11180/) (2019)
- [Benchmarking Multi-Agent Deep Reinforcement Learning Algorithms in Cooperative Tasks](https://mlsanity.com/paper/2006.07869/) (2020)
- [RUDDER: Return Decomposition for Delayed Rewards](https://mlsanity.com/paper/1806.07857/) (2018)
- [PRIMAL: Pathfinding via Reinforcement and Imitation Multi-Agent Learning](https://mlsanity.com/paper/1809.03531/) (2018)
- [Hindsight Credit Assignment](https://mlsanity.com/paper/1912.02503/) (2019)
- [Shapley Q-value: A Local Reward Approach to Solve Global Reward Games](https://mlsanity.com/paper/1907.05707/) (2019)
- [Inverse Dynamics Pretraining Learns Good Representations for Multitask Imitation](https://mlsanity.com/paper/2305.16985/) (2023)
- [Learning Guidance Rewards with Trajectory-space Smoothing](https://mlsanity.com/paper/2010.12718/) (2020)
- [Scalable Multi-Agent Reinforcement Learning for Warehouse Logistics with Robotic and Human Co-Workers](https://mlsanity.com/paper/2212.11498/) (2022)
- [Sequence Modeling of Temporal Credit Assignment for Episodic Reinforcement Learning](https://mlsanity.com/paper/1905.13420/) (2019)
- [Decoupled Reinforcement Learning to Stabilise Intrinsically-Motivated Exploration](https://mlsanity.com/paper/2107.08966/) (2021)
- [Agent-Temporal Attention for Reward Redistribution in Episodic Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2201.04612/) (2022)
- [SMAClite: A Lightweight Environment for Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2305.05566/) (2023)
- [Learning Cooperative Multi-Agent Policies with Partial Reward Decoupling](https://mlsanity.com/paper/2112.12740/) (2021)
- [STAS: Spatial-Temporal Return Decomposition for Multi-agent Reinforcement Learning](https://mlsanity.com/paper/2304.07520/) (2023)
- [Assigning Credit with Partial Reward Decoupling in Multi-Agent Proximal Policy Optimization](https://mlsanity.com/paper/2408.04295/) (2024)

## Abstract (arXiv)

Credit assignmen, disentangling each agent's contribution to a shared reward, is a critical challenge in cooperative multi-agent reinforcement learning (MARL). To be effective, credit assignment methods must preserve the environment's optimal policy. Some recent approaches attempt this by enforcing return equivalence, where the sum of distributed rewards must equal the team reward. However, their guarantees are conditional on a learned model's regression accuracy, making them unreliable in practice. We introduce Temporal-Agent Reward Redistribution (TAR2), an approach that decouples credit modeling from this constraint. A neural network learns unnormalized contribution scores, while a separate, deterministic normalization step enforces return equivalence by construction. We demonstrate that this method is equivalent to a valid Potential-Based Reward Shaping (PBRS), which guarantees the optimal policy is preserved regardless of model accuracy. Empirically, on challenging SMACLite and Google Research Football (GRF) benchmarks, TAR2 accelerates learning and achieves higher final performance than strong baselines. These results establish our method as an effective solution for the agent-temporal credit assignment problem.