Research map / Reinforcement learning / Visual and representation RL
RAST-MoE-RL: A Regime-Aware Spatio-Temporal MoE Framework for Deep Reinforcement Learning in Ride-Hailing
Yuhan Tang, Kangxin Cui, Jung Ho Park, Yibo Zhao, Xuan Jiang, Haoze He, Jiangbo Yu, Haris Koutsopoulos, Jinhua Zhao
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 |
|---|---|---|---|
| replace the 12M plain MLP encoder with the 12M MoE (16,4) encoder under PPO (ablation) | PPO policy · SF Uber trajectories (held-out test) | matching wait (s): 208 -> 182 | Improved |
| replace the 12M plain MLP encoder with the 12M MoE (16,4) encoder under PPO (ablation) | PPO policy · SF Uber trajectories (held-out test) | pickup wait (s): 577 -> 523 | Improved |
| replace a 48M dense Transformer with a 12M MoE (16,4) encoder (ablation) | PPO policy · SF Uber trajectories (held-out test) | pickup wait (s): 545 -> 523 | Improved |
| replace a 12M dense Transformer with a 12M MoE (16,4) encoder (ablation) | PPO policy · SF Uber trajectories (held-out test) | matching wait (s): 201 -> 182 | Improved |
| use top-4 routing instead of top-2 with 16 experts (ablation) | MoE PPO · SF Uber trajectories (held-out test) | pickup wait (s): 543 -> 523 | Improved |
| scale to 32 experts (top-8, 24M) from 16 experts (top-4, 12M) (ablation) | MoE PPO · SF Uber trajectories (held-out test) | matching wait (s): 182 -> 188 | Hurt |
| scale to 64 experts (top-8, 48M) from 16 experts (top-4, 12M) (ablation) | MoE PPO · SF Uber trajectories (held-out test) | pickup wait (s): 523 -> 536 | Hurt |
| use 8 experts top-4 (6M) instead of 16 experts top-4 (ablation) | MoE PPO · SF Uber trajectories (held-out test) | pickup wait (s): 523 -> 541 | Hurt |
| use GRPO-style within-state normalization instead of PPO (ablation) | MoE (16,4) · SF Uber trajectories (held-out test) | pickup wait (s): 523 -> 528 | No clear effect |
| use PPO instead of A2C (ablation) | MoE (16,4) · SF Uber trajectories (held-out test) | pickup wait (s): 555 -> 523 | Improved |
| use PPO instead of DQN (ablation) | MoE (16,4) · SF Uber trajectories (held-out test) | pickup wait (s): 612 -> 523 | Improved |
| use adaptive reward instead of fixed ratio 8:1 (ablation) | PPO 16 experts top-4 · SF Uber trajectories (reward ratio, PPO 16 experts top-4) | pickup wait (s): 1746 -> 524 | Improved |
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)
- NondBREM: Nondeterministic Offline Reinforcement Learning for Large-Scale Order Dispatching (AAAI 2024 · similarity 0.86)
- GARLIC: GPT-Augmented Reinforcement Learning with Intelligent Control for Vehicle Dispatching (AAAI 2025 · similarity 0.85)
- Hierarchical Optimization via LLM-Guided Objective Evolution for Mobility-on-Demand Systems (NeurIPS 2025 · similarity 0.85)
- i-Rebalance: Personalized Vehicle Repositioning for Supply Demand Balance (AAAI 2024 · similarity 0.84)
- Massively Scalable Inverse Reinforcement Learning in Google Maps (ICLR 2024 · similarity 0.83)
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.
