Research map / Multi-agent systems and games / Cooperative multi-agent RL
Multi-Agent Pointer Transformer: Seq-to-Seq Reinforcement Learning for Multi-Vehicle Dynamic Pickup-Delivery Problems
Zengyu Zou, Jingyuan Wang, Yixuan Huang, Junjie Wu
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 |
|---|---|---|---|
| add autoregressive joint decoding (ablation) | MAPT · synth-XL | Obj.: 1905.8 -> 3227.5 | Improved |
| add autoregressive joint decoding (ablation) | MAPT · dhrd-se | Obj.: 49.4 -> 78.9 | Improved |
| add informative priors (load balance, distance) (ablation) | MAPT · synth-XL | Obj.: 1375.0 -> 3227.5 | Improved |
| add informative priors (ablation) | MAPT · synth-S | Obj.: 175.2 -> 275.2 | Improved |
| add Relation-Aware Attention (ablation) | MAPT · synth-S | Obj.: 270.9 -> 275.2 | Improved |
| add Relation-Aware Attention (ablation) | MAPT · dhrd-se | Obj.: 77.1 -> 78.9 | Improved |
| use MAPT instead of MAPDP (independent decoding, no priors) | MAPT · synth-S | Obj.: 157.5 -> 275.2 | Improved |
| use MAPT instead of Rolling-Horizon SA | MAPT · synth-L inference time | seconds per instance: 166.660 -> 8.300 | Improved |
| use MAPT instead of Nearest rule | MAPT · dhrd-se | Obj.: 39.8 -> 78.9 | Improved |
| increase deferral prior beta to larger values (ablation) | MAPT · synth-L, synth-XL, dhrd-tpe, dhrd-se | Obj. | Hurt |
| use small non-zero beta (ablation) | MAPT · dhrd-tpe, dhrd-se | Obj. | Improved |
| MAPT versus Nearest on inference time | MAPT · synth-S | seconds per instance: 0.059 -> 1.420 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MAPDP — MAPDP: Cooperative Multi-Agent Reinforcement Learning to Solve Pickup and Delivery Problems (AAAI 2022) · better on 5 of 5 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- One Step is Enough: Multi-Agent Reinforcement Learning Based on One-Step Policy Optimization for Order Dispatch on Ride-Sharing Platforms (NeurIPS 2026 · similarity 0.87)
- DriveGPT: Scaling Autoregressive Behavior Models for Driving (ICML 2025 · similarity 0.85)
- MTL-KD: Multi-Task Learning Via Knowledge Distillation for Generalizable Neural Vehicle Routing Solver (NeurIPS 2025 · similarity 0.85)
- MAPF-GPT: Imitation Learning for Multi-Agent Pathfinding at Scale (AAAI 2025 · similarity 0.85)
- BehaviorGPT: Smart Agent Simulation for Autonomous Driving with Next-Patch Prediction (NeurIPS 2024 · similarity 0.84)
- Decision ConvFormer: Local Filtering in MetaFormer is Sufficient for Decision Making (ICLR 2024 · 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.
