# QAMM: Adjoint MeanFlow Matching for Few-Step Offline Reinforcement Learning

arXiv: https://arxiv.org/abs/2609.34497 · HTML: https://mlsanity.com/paper/2609.34497/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Diffusion and flow policies for robotics, one of 14 directions in Diffusion models.
- Builds on 23 of its 26 references that are in the map.
- First posted on arXiv 28 Sep 2026.

## Placed in

- Diffusion models / [Diffusion and flow policies for robotics](https://mlsanity.com/diffusion/diffusion-and-flow-policies-for-robotics/)
- Offline RL and imitation learning from data / [Diffusion and flow policies for offline RL](https://mlsanity.com/offlinerl/diffusion-and-flow-policies-for-offline-rl/)

## Builds on (most cited first, 23 of 23)

- [Flow Matching for Generative Modeling](https://mlsanity.com/paper/2210.02747/) (2022)
- [Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow](https://mlsanity.com/paper/2209.03003/) (2022)
- [Offline Reinforcement Learning with Implicit Q-Learning](https://mlsanity.com/paper/2110.06169/) (2021)
- [Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning](https://mlsanity.com/paper/2208.06193/) (2022)
- [Mean Flows for One-step Generative Modeling](https://mlsanity.com/paper/2505.13447/) (2025)
- [IDQL: Implicit Q-Learning as an Actor-Critic Method with Diffusion Policies](https://mlsanity.com/paper/2304.10573/) (2023)
- [OGBench: Benchmarking Offline Goal-Conditioned RL](https://mlsanity.com/paper/2410.20092/) (2024)
- [Adjoint Matching: Fine-tuning Flow and Diffusion Generative Models with Memoryless Stochastic Optimal Control](https://mlsanity.com/paper/2409.08861/) (2024)
- [Flow Q-Learning](https://mlsanity.com/paper/2502.02538/) (2025)
- [Steering Your Diffusion Policy with Latent Space Reinforcement Learning](https://mlsanity.com/paper/2506.15799/) (2025)
- [Learning a Diffusion Model Policy from Rewards via Q-Score Matching](https://mlsanity.com/paper/2312.11752/) (2023)
- [Revisiting the Minimalist Approach to Offline Reinforcement Learning](https://mlsanity.com/paper/2305.09836/) (2023)
- [Diffusion-based Reinforcement Learning via Q-weighted Variational Policy Optimization](https://mlsanity.com/paper/2405.16173/) (2024)
- [Improved Mean Flows: On the Challenges of Fastforward Generative Models](https://mlsanity.com/paper/2512.02012/) (2025)
- [Consistency Models as a Rich and Efficient Policy Class for Reinforcement Learning](https://mlsanity.com/paper/2309.16984/) (2023)
- [Energy-Weighted Flow Matching for Offline Reinforcement Learning](https://mlsanity.com/paper/2503.04975/) (2025)
- [Much Ado About Noising: Dispelling the Myths of Generative Robotic Control](https://mlsanity.com/paper/2512.01809/) (2025)
- [Diffusion Actor-Critic: Formulating Constrained Policy Iteration as Diffusion Noise Regression for Offline Reinforcement Learning](https://mlsanity.com/paper/2405.20555/) (2024)
- [Scaling Offline RL via Efficient and Expressive Shortcut Models](https://mlsanity.com/paper/2505.22866/) (2025)
- [One-Step Flow Policy Mirror Descent](https://mlsanity.com/paper/2507.23675/) (2025)
- [FlowQ: Energy-Guided Flow Policies for Offline Reinforcement Learning](https://mlsanity.com/paper/2505.14139/) (2025)
- [MeanFlowNFT: Bringing Forward-Process RL to Average-Velocity Generators](https://mlsanity.com/paper/2607.15273/) (2026)
- [Trust Region Q Adjoint Matching](https://mlsanity.com/paper/2605.27079/) (2026)

## Abstract (arXiv)

Flow policies can model rich action distributions, but their iterative sampling limits decision speed. Adjoint matching uses the critic's action gradient to improve a flow policy without backpropagating through its sampling trajectory, yet its supervision is defined for instantaneous velocities. We propose QAMM, a method that turns the critic-derived adjoint signal into supervision for MeanFlow's average velocity. The resulting policy learns finite-interval transport directly and generates actions with few network evaluations. We derive the adjoint MeanFlow target, specify its gradient boundaries, and train it with an offline actor-critic. On ten HumanoidMaze tasks, QAMM produces effective two-call policies and achieves competitive performance against strong flow-policy baselines. These results show that adjoint-based Q optimization can be combined with average-velocity learning to obtain expressive offline policies with few-step action generation.