Guided Flow Policy:
Learning from High-Value Actions in Offline Reinforcement Learning

ArXiv    HAL    Webpage

Guided Flow Policy (GFP) is an offline RL method based on flow matching. It couples a multi-step flow-matching policy trained with value-aware behavior cloning and a distilled one-step actor through a bidirectional guidance mechanism. This enables GFP to achieve state-of-the-art performance across 144 state and pixel-based tasks from the OGBench, Minari, and D4RL benchmarks, with substantial gains on suboptimal datasets and challenging tasks.

News & Updates

• 🟢 2025-12-03 - Release of the paper on ArXiv
• 🔴 Code, coming soon
• 🔴 Detailed blog post, coming soon