Research map / Multi-agent systems and games / Cooperative multi-agent RL
Multi-agent active perception with prediction rewards
Mikko Lauri, Frans Oliehoek
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 adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · MAV h=2 | policy value: -2.112 -> -2.006 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · MAV h=3 | policy value: -2.002 -> -1.936 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · MAV h=4 | policy value: -1.943 -> -1.879 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · MAV h=5 | policy value: -1.918 -> -1.842 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · Rovers h=2 | policy value: -3.800 -> -3.484 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · Rovers h=3 | policy value: -3.680 -> -3.402 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · Rovers h=4 | policy value: -3.640 -> -3.367 | Improved |
| Add adaptive α-vector selection rather than randomly resampling α-vectors (ablation) | APAS · Rovers h=5 | policy value: -3.591 -> -3.293 | Improved |
| Use APAS instead of NPGI | MAV · MAV h=2 | policy value: -1.931 -> -2.006 | Hurt |
| Use APAS instead of NPGI | Rovers · Rovers h=2 | policy value: -3.495 -> -3.484 | Improved |
| Extend planning beyond horizon 5 using APAS rather than NPGI | Multi-agent active perception planning · MAV | maximum completed planning horizon: h>5 | Improved |
| Increase the number of α-vectors from K=5 to K=10 (ablation) | APAS · Rovers h=10 | policy value | No clear effect |
Closest papers by meaning
- Efficient Multiagent Planning via Shared Action Suggestions (AAAI 2026 · similarity 0.88)
- ACTIVE-o3 : Empowering MLLMs with Active Perception via Pure Reinforcement Learning (ICML 2026 · similarity 0.85)
- Factored Online Planning in Many-Agent POMDPs (AAAI 2024 · similarity 0.85)
- Rainbow Delay Compensation: A Multi-Agent Reinforcement Learning Framework for Mitigating Observation Delays (NeurIPS 2025 · similarity 0.85)
- An Extensible Framework for Open Heterogeneous Collaborative Perception (ICLR 2024 · similarity 0.84)
- Real-World Reinforcement Learning of Active Perception Behaviors (NeurIPS 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.
