Conditioning Matters: Training Diffusion Policies is Faster Than You Think

• URL: http://arxiv.org/abs/2505.11123v1
• Date: Fri, 16 May 2025 11:14:22 GMT
• Title: Conditioning Matters: Training Diffusion Policies is Faster Than You Think
• Authors: Zibin Dong, Yicheng Liu, Yinchuan Li, Hang Zhao, Jianye Hao,
• Abstract summary: Diffusion policies have emerged as a mainstream paradigm for building vision-language-action (VLA) models.<n>We identify a fundamental challenge in conditional diffusion policy training: when generative conditions are hard to distinguish, the training objective degenerates into modeling the marginal action distribution.<n>We propose Cocos, a solution that modifies the source distribution in the conditional flow matching to be condition-dependent.
• Score: 69.31534053485711
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Diffusion policies have emerged as a mainstream paradigm for building vision-language-action (VLA) models. Although they demonstrate strong robot control capabilities, their training efficiency remains suboptimal. In this work, we identify a fundamental challenge in conditional diffusion policy training: when generative conditions are hard to distinguish, the training objective degenerates into modeling the marginal action distribution, a phenomenon we term loss collapse. To overcome this, we propose Cocos, a simple yet general solution that modifies the source distribution in the conditional flow matching to be condition-dependent. By anchoring the source distribution around semantics extracted from condition inputs, Cocos encourages stronger condition integration and prevents the loss collapse. We provide theoretical justification and extensive empirical results across simulation and real-world benchmarks. Our method achieves faster convergence and higher success rates than existing approaches, matching the performance of large-scale pre-trained VLAs using significantly fewer gradient steps and parameters. Cocos is lightweight, easy to implement, and compatible with diverse policy architectures, offering a general-purpose improvement to diffusion policy training.

Related papers

• VARD: Efficient and Dense Fine-Tuning for Diffusion Models with Value-based RL [28.95582264086289]
  VAlue-based Reinforced Diffusion (VARD) is a novel approach that first learns a value function predicting expection of rewards from intermediate states.<n>Our method maintains proximity to the pretrained model while enabling effective and stable training via backpropagation.
  arXiv  Detail & Related papers  (2025-05-21T17:44:37Z)
• Diffusion Actor-Critic: Formulating Constrained Policy Iteration as Diffusion Noise Regression for Offline Reinforcement Learning [13.163511229897667]
  In offline reinforcement learning, it is necessary to manage out-of-distribution actions to prevent overestimation of value functions.<n>We propose Diffusion Actor-Critic (DAC) that formulates the Kullback-Leibler (KL) constraint policy iteration as a diffusion noise regression problem.<n>Our approach is evaluated on D4RL benchmarks and outperforms the state-of-the-art in nearly all environments.
  arXiv  Detail & Related papers  (2024-05-31T00:41:04Z)
• Diffusion-based Reinforcement Learning via Q-weighted Variational Policy Optimization [55.97310586039358]
  Diffusion models have garnered widespread attention in Reinforcement Learning (RL) for their powerful expressiveness and multimodality.<n>We propose a novel model-free diffusion-based online RL algorithm, Q-weighted Variational Policy Optimization (QVPO)<n>Specifically, we introduce the Q-weighted variational loss, which can be proved to be a tight lower bound of the policy objective in online RL under certain conditions.<n>We also develop an efficient behavior policy to enhance sample efficiency by reducing the variance of the diffusion policy during online interactions.
  arXiv  Detail & Related papers  (2024-05-25T10:45:46Z)
• Belief-Enriched Pessimistic Q-Learning against Adversarial State Perturbations [5.076419064097735]
  Recent work shows that a well-trained RL agent can be easily manipulated by strategically perturbing its state observations at the test stage. Existing solutions either introduce a regularization term to improve the smoothness of the trained policy against perturbations or alternatively train the agent's policy and the attacker's policy. We propose a new robust RL algorithm for deriving a pessimistic policy to safeguard against an agent's uncertainty about true states.
  arXiv  Detail & Related papers  (2024-03-06T20:52:49Z)
• Don't Start from Scratch: Behavioral Refinement via Interpolant-based Policy Diffusion [16.44141792109178]
  Diffusion models learn to shape a policy by diffusing actions (or states) from standard Gaussian noise. The target policy to be learned is often significantly different from Gaussian and this can result in poor performance when using a small number of diffusion steps. We contribute both theoretical results, a new method, and empirical findings that show the benefits of using an informative source policy.
  arXiv  Detail & Related papers  (2024-02-25T12:19:21Z)
• Learning a Diffusion Model Policy from Rewards via Q-Score Matching [93.0191910132874]
  We present a theoretical framework linking the structure of diffusion model policies to a learned Q-function.<n>We propose a new policy update method from this theory, which we denote Q-score matching.
  arXiv  Detail & Related papers  (2023-12-18T23:31:01Z)
• Exploiting Diffusion Prior for Real-World Image Super-Resolution [75.5898357277047]
  We present a novel approach to leverage prior knowledge encapsulated in pre-trained text-to-image diffusion models for blind super-resolution. By employing our time-aware encoder, we can achieve promising restoration results without altering the pre-trained synthesis model.
  arXiv  Detail & Related papers  (2023-05-11T17:55:25Z)
• Diffusion Policies as an Expressive Policy Class for Offline Reinforcement Learning [70.20191211010847]
  Offline reinforcement learning (RL) aims to learn an optimal policy using a previously collected static dataset. We introduce Diffusion Q-learning (Diffusion-QL) that utilizes a conditional diffusion model to represent the policy. We show that our method can achieve state-of-the-art performance on the majority of the D4RL benchmark tasks.
  arXiv  Detail & Related papers  (2022-08-12T09:54:11Z)
• Re-parameterizing VAEs for stability [1.90365714903665]
  We propose a theoretical approach towards the training numerical stability of Variational AutoEncoders (VAE) Our work is motivated by recent studies empowering VAEs to reach state of the art generative results on complex image datasets. We show that by implementing small changes to the way we parameterize the Normal distributions on which they rely, VAEs can securely be trained.
  arXiv  Detail & Related papers  (2021-06-25T16:19:09Z)
• Strictly Batch Imitation Learning by Energy-based Distribution Matching [104.33286163090179]
  Consider learning a policy purely on the basis of demonstrated behavior -- that is, with no access to reinforcement signals, no knowledge of transition dynamics, and no further interaction with the environment. One solution is simply to retrofit existing algorithms for apprenticeship learning to work in the offline setting. But such an approach leans heavily on off-policy evaluation or offline model estimation, and can be indirect and inefficient. We argue that a good solution should be able to explicitly parameterize a policy, implicitly learn from rollout dynamics, and operate in an entirely offline fashion.
  arXiv  Detail & Related papers  (2020-06-25T03:27:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.