Flow-based vision-language-action (VLA) models achieve strong performance in robotics through imitation learning, yet fine-tuning them with on-policy reinforcement learning (RL) is limited. In group-based RL methods like Group Relative Policy Optimization (GRPO), every rollout step requires multi-step ODE integration, and the sampler's limited stochasticity leads to rollout groups with low variance. We propose VLA-SSD, which addresses both high inference latency and limited stochasticity. In particular, our method self-distills a pretrained flow-matching action head into a stochastic one-step policy by conditioning a single-jump generator on a Brownian path. We derive a conditional action density via our stochastic one-step policy, enabling efficient post-training via likelihood-ratio RL algorithms. We show that distilling VLA-SSD from the π0 VLA reduces the action head's forward-pass time by a factor of 7.7, at the cost of a 5.5% relative decrease in the success rate on the LIBERO benchmark. However, by leveraging the increased sample diversity of VLA-SSD during group generation, post-training a π0.5 VLA using GRPO yields a 20.6% relative improvement.
Figures & tables
Figure 1: Left: an ODE sampler and Flow-SDE spend K action-expert evaluations per chunk; VLA-SSD spends one and keeps the spread of the SDE. Right: for a fixed s and a0 , the Itô flow map conditions on a Brownian path W(k) , so G paths give G different one-step actions with different returns, and the GRPO group carries gradient; at σ=0 all G coincide.
Policy
Sampler
Passes
Forward (ms)
LIBERO
MetaWorld
π0.5 SFT
ODE, K=10
10
58.4
77.1
51.8
πRL , GRPO
SDE, K=4
4
24.5
91.5
—
πRL , PPO
SDE, K=4
4
24.5
98.3
70.7
Itô student (ours)
Itô, K=1
1
7.6
68.6
49.0
Itô + GRPO (ours)
Itô, K=1
1
7.6
93.0
72.0
Table 1: Main results. Success on the LIBERO four-suite mean and MetaWorld MT50 (ours at σ=0 ); passes is action-expert evaluations per chunk, forward its time per chunk on one A100. πRL rows are its published numbers from the same SFT checkpoints. Bold is best per column.
Map
Objective
Forward (ms)
spatial
object
goal
long
mean
π0 teacher, K=10
—
58.4
96.8
98.8
95.8
85.2
94.15
Deterministic flow map
LSD
7.6
96.2
96.0
94.0
79.0
91.30
PSD
7.6
96.4
95.4
93.6
77.8
90.80
ESD
7.6
90.2
76.8
77.0
26.0
67.50
Itô flow map, σ=0.1
LSD
7.6
94.2
93.6
91.0
77.0
88.95
PSD
7.6
96.6
91.2
90.8
73.2
87.95
Table 4: One-step distillation from π0 on LIBERO, 500 episodes per suite, all students evaluated at K=1 . The teacher is πRL ’s full-dataset SFT π0 checkpoint at K=10 ; the deterministic map under ESD is MeanFlow ( Geng et al., 2025 ) . Forward is the action head’s time per chunk on one A100. Bold marks the best objective for each map; the standard error on each mean is about 0.7 points.
Appendix figures & tables15 assets
Supplementary material from the paper’s appendix.
Appendix
RL checkpoint
Mean absolute log-ratio error (FP32)
Early
0.000181
Middle
0.000070
Late
0.000061
Appendix
Table 5: FP32 log-ratio error across GRPO checkpoints.
Figure 3: LIBERO-Spatial , task 4. Pick up the black bowl in the top drawer of the wooden cabinet and place it on the plate.
Figure 4: LIBERO-Spatial , task 7. Pick up the black bowl on the stove and place it on the plate.
Figure 5: LIBERO-Object , task 0. Pick up the alphabet soup and place it in the basket.
Figure 6: LIBERO-Object , task 7. Pick up the milk and place it in the basket.
Figure 7: LIBERO-Goal , task 0. Open the middle drawer of the cabinet.
Figure 8: LIBERO-Goal , task 5. Push the plate to the front of the stove.
Figure 9: LIBERO-Long , task 2. Turn on the stove and put the moka pot on it.
Figure 10: LIBERO-Long , task 4. Put the white mug on the left plate and put the yellow and white mug on the right plate.
Figure 11: assembly-v3 (MetaWorld MT50). Pick up a nut and place it onto a peg.
Figure 12: basketball-v3 (MetaWorld MT50). Dunk the basketball into the basket.
Figure 13: button-press-topdown-v3 (MetaWorld MT50). Press a button from the top.
Figure 14: drawer-open-v3 (MetaWorld MT50). Open a drawer.
Figure 15: shelf-place-v3 (MetaWorld MT50). Pick and place a puck onto a shelf.
Figure 16: window-open-v3 (MetaWorld MT50). Push and open a window.
Vision-language-action (VLA) models with flow-matching action heads, such as π0 and π0.5, have become a standard recipe for generalist robot manipulation, but every action chunk is produced by integrating a learned velocity field over about ten sequential action-expert evaluations, so action generation dominates model latency and caps the achievable control frequency. Existing remedies either shrink each evaluation through layer distillation or pruning while leaving the sequential solver in place, or import one-step objectives from image generation that were designed for training from scratch, rely on an exponential-moving-average teacher or an external critic, and regress finite-span targets onto the conditional velocity, whose variance around the marginal field biases the learned map. We introduce SnapFlow, a self-distillation method that adapts a pretrained flow-matching VLA to one-step action generation. SnapFlow mixes local flow-matching supervision with shortcut samples whose target is the model's own two-step Euler endpoint under stop-gradient, and adds a single zero-initialized destination-time projection so that one action expert serves both objectives while the pretrained predictor is preserved at initialization; no separate teacher network or critic is required. On π0.5, whose 10-NFE teacher operates near ceiling, the archived one-NFE SnapFlow checkpoint reaches 98.75% LIBERO success versus 97.75% for the teacher; under a separate paired evaluation of π0, the one-NFE model surpasses its teacher by 10.5 percentage points. SnapFlow also lowers mean and tail action error on held-out data, cuts batch-one model latency by about 3.3x, and transfers to shelf retrieval on a dual-arm humanoid.
Wuyang Luan, Junhui Li, Weiguang Zhao +2
Jilin University · Chongqing University · University of Liverpool +1
Vision-language-action (VLA) models increasingly rely on diffusion- or flow-matching-based action heads to generate continuous robot actions. These action heads typically process the denoising trajectory in a largely uniform manner. However, we observe that the conditioning focus naturally shifts across denoising stages: early stages combine language instructions and visual observations to establish a coarse action trajectory, whereas later stages place greater emphasis on current visual observations for action alignment. Based on this insight, we introduce StairVLA, a stage-aware hierarchical action generation framework that uses partially denoised actions as a natural interface between coarse long-horizon action generation and local refinement. A high-level VLA performs early denoising to produce a reusable long-horizon partially denoised action trajectory, while a lightweight refiner operates at a higher frequency to refine local action chunks using the latest observations. This design amortizes expensive high-level VLA computation while preserving frequent closed-loop correction. On LIBERO, our GR00T-style instantiation improves average success from 96.5% to 97.8% while reducing amortized inference latency from 115.0 ms to 44.2 ms per action chunk. More broadly, across two VLA backbones, simulation benchmarks, and real-robot tasks, StairVLA consistently reduces inference cost while maintaining strong task performance.
Shangyuan Yuan, Xinda Qi, Yujiang Pu +2
Michigan State University · Ant Technology U.S., Inc.
Specialized vision-language-action (VLA) policies acquire strong domain-specific skills, yet consolidating these skills into one shared policy remains challenging. On-policy distillation (OPD) transfers teacher knowledge on inputs induced by the student, addressing the distribution mismatch encountered during deployment. For flow-based VLAs, this mismatch arises at two levels: environment observations and intermediate action-generation states. We propose a two-level on-policy distillation framework that couples student-driven environment interaction with teacher velocity supervision along student-generated action flow trajectories. Frozen domain specialists jointly supervise one student. After each update, the student collects new environment experience and generates fresh action-flow trajectories. Starting from a joint supervised fine-tuning (SFT) initialization, Velocity Loss OPD yields absolute success-rate gains of 22.55% on LIBERO-40, 33.4% on CALVIN five-instruction sequences, and 22.6% on Meta-World MT50. Compared with Action Loss OPD and Teacher-rollout SFT, it achieves absolute success-rate gains of 6.4-9.5% on LIBERO-40, 9.7-18.3% on CALVIN, and 3.2-4.2% on Meta-World MT50. Analysis by sequence length further shows larger success-rate gains on longer CALVIN sequences. These results demonstrate effective multi-domain capability consolidation through two-level on-policy supervision, with a single student deployed across domains.
Lichao Ni, Yuhan Liu, Pengfei Li +1
Huawei Technologies Co., Ltd. · Tsinghua University