Research map / Reinforcement learning / MDPs and planning
Efficient Multiagent Planning via Shared Action Suggestions
Dylan M. Asmar, Mykel J. Kochenderfer
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 |
|---|---|---|---|
| use alpha vector indices for belief subspace refinement instead of action only (ablation) | MCAS · Meet 27 UI, WP | reward: 1.6 -> 2.0 | Improved |
| use alpha vector indices instead of action only (ablation) | MCAS · Mars Rover UI (2 agents) | reward: 19.8 -> 23.9 | Improved |
| use alpha vector indices instead of action only (ablation) | MCAS · Mars Rover 5G, UI | reward: 18.0 -> 20.7 | Improved |
| use alpha vector indices instead of action only (ablation) | MCAS · Box Push SO | reward: 199.8 -> 203.2 | Improved |
| use alpha vector indices instead of action only (ablation) | MCAS · Dec-Tiger (2 agents) | reward: 58.5 -> 58.5 | No clear effect |
| alpha vector indices shrink max belief set size (ablation) | MCAS · Box Push SO | max belief set size: 192.1 -> 4.8 | Improved |
| alpha vector indices shrink max belief set size (ablation) | MCAS · Meet 27 UI, WP | max belief set size: 16.8 -> 1.5 | Improved |
| share observations (MPOMDP) rather than control with individual observations only | MPOMDP-I · Meet 3x3 AG, UI, WP (3 agents) | reward: 1.0 -> 6.4 | Improved |
| share observations (MPOMDP) rather than individual observations only | MPOMDP-I · Wireless (2 agents) | reward: -152.8 -> -152.8 | No clear effect |
| share observations (MPOMDP) rather than individual only | MPOMDP-I · Broadcast (2 agents) | reward: 9.4 -> 9.4 | No clear effect |
| conflate true individual beliefs (MPOMDP-C) instead of centralized belief | MPOMDP · Dec-Tiger (2 agents) | reward: 59.5 -> 59.5 | No clear effect |
| MCAS action-suggestion communication versus independent execution | Independent · Dec-Tiger (2 agents) | reward: -68.1 -> 58.5 | Improved |
Closest papers by meaning
- Knowledge-Based Policies for Qualitative Decentralized POMDPs (AAAI 2018 · similarity 0.89)
- Factored Online Planning in Many-Agent POMDPs (AAAI 2024 · similarity 0.89)
- Belief-State Query Policies for User-Aligned POMDPs (NeurIPS 2024 · similarity 0.87)
- Optimally Solving Simultaneous-Move Dec-POMDPs: The Sequential Central Planning Approach (AAAI 2025 · similarity 0.87)
- Multi-Environment POMDPs with Finite-Horizon Objectives (NeurIPS 2026 · similarity 0.87)
- Revelations: A Decidable Class of POMDPs with Omega-Regular Objectives (AAAI 2025 · 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.
