Staircase Policy: Streaming Inference for World-Action Models with Large Action Chunks

Guoheng Sun; Chen Chen; Jin Wang; Ang Li; Teresa Lv

cs.ROSep 29, 2026

Staircase Policy: Streaming Inference for World-Action Models with Large Action Chunks

Authors: Guoheng Sun, Chen Chen, Jin Wang, Ang Li, Teresa Lv

Organizations: University of Maryland, College Park · Independent Researcher · Oxford Robotics Institute, University of Oxford

Abstract

World-Action Models (WAMs) improve robotic manipulation by conditioning action generation on predicted future observations, but future prediction adds further inference overhead to already expensive iterative action generation. Action chunking can amortize this cost over multiple actions, yet performance degrades over long execution horizons because later actions remain conditioned on stale observations. We introduce STAIRCASE POLICY, a streaming inference and training framework that turns a flow-matching VLA into a JEPA-style WAM and partitions a large action chunk into sub-chunks at staggered denoising stages. Near-term actions are executed as soon as they become available, while later actions continue to be refined. At each sub-chunk boundary, the future latent is re-predicted from the latest observation and used to update all unexecuted actions, enabling long-horizon execution without repeated full policy inference. The resulting future-prediction error can further serve as a signal for adaptive chunking. S-WAM achieves 97.7% on LIBERO and 87.9% on LIBERO-Plus, and improves performance across multiple policy backbones and real-robot tasks. It reaches 292.7 executed actions per second, $3.62\times$ the throughput of conventional execution at comparable accuracy, while reducing time-to-first-action from 123.6 to 73.3 ms. With additional inference optimizations, throughput further increases to 642.9 actions per second.

Figures & tables

Figure 1: Success rate versus measured inference throughput on LIBERO. The measurement protocol, per-method results, and methods omitted from the figure are provided in Appendix C .

Figure 2: Overview of Staircase Policy . Here $t_{i}$ denotes a time step and $t_{i}^{*}$ the predicted future.

Figure 3: Real-robot results on nine tasks. Task settings and details are provided in Appendix F .

Method	Background	Robot	Camera	Language	Noise	Layout	Light	Avg.
ST-WAM ( Wang et al., 2026b )	74.2	60.1	55.4	79.3	79.5	74.3	93.0	73.7
DreamWAM ( Yuan et al., 2026a )	71.5	63.6	53.7	94.8	67.1	80.7	96.6	75.4
VLA-JEPA ( Sun et al., 2026c )	93.6	67.1	63.3	85.4	66.3	85.1	95.6	79.5
ROCKET-VLA ( Sun et al., 2026a )	91.8	41.8	91.8	78.0	92.5	81.2	94.7	81.7
VLANeXt ( Wu et al., 2026 )	82.5	65.7	90.4	81.8	94.1	80.8	95.9	84.5
WorldPilot ( Lin et al., 2026c )	96.4	60.6	82.8	87.2	93.6	80.5	98.6	85.7

Table 1: Robustness on LIBERO-Plus across seven perturbation dimensions.

Figure 4: Success rate versus training steps across five global batch sizes.

Figure 6: Success rate versus execution horizon, averaged over four suites.

Figure 7: Comparison with compact VLA models on LIBERO, LIBERO-Plus, DOMINO.

Figure 8: SR against throughput for fixed and adaptive commitment on LIBERO-Long.

Backbone	Policy	Spatial	Object	Goal	Long	Avg.	Act/s $\uparrow$	TTFA $\downarrow$
		Success rate					Speed
LaWAM	S-WAM	98.6	100.0	97.0	95.0	97.65	292.7	73.3
( Chen et al., 2026a )	Vanilla, $H_{\mathrm{exec}}{=}10$	96.2	98.4	95.6	90.6	95.20	80.9	123.6
(2.3B)	Vanilla, $H_{\mathrm{exec}}{=}50$	87.2	80.6	88.0	73.8	82.40	402.5	124.2
$\pi_{0.5}$	S-WAM	98.0	98.6	95.2	94.2	96.50	243.2	80.6
( Intelligence et al., 2025 )	Vanilla, $H_{\mathrm{exec}}{=}10$	94.6	99.0	91.6	87.4	93.15	75.7	132.2

Table 3: Staircase Policy transfers across different backbones. Act/s is executed actions per second; TTFA is time to first action (ms).

Appendix figures & tables18 assets

Supplementary material from the paper’s appendix.

Appendix

	LIBERO	LIBERO-Plus	DOMINO
Cosine horizon	25k	25k	50k
S-WAM , steps trained	10k	10k	20k
S-WAM , reported checkpoint	5k	8k	20k
Vanilla, steps trained	25k	25k	50k
Vanilla, reported checkpoint	20k	25k	50k

Appendix

Table 4: Training-step budgets for the LaWAM host.

	Training FLOPs (EFLOPs)			GPU-hours
	S-WAM	Vanilla	Ratio	S-WAM	Vanilla	Ratio
LIBERO	3.75	5.02	$0.75\times$	11.4	12.9	$0.89\times$
LIBERO-Plus	3.75	5.02	$0.75\times$	11.6	12.8	$0.90\times$
DOMINO	8.74	12.99	$0.67\times$	37.0	47.4	$0.78\times$

Appendix

Table 5: Total training cost on the LaWAM backbone.

	$\pi_{0.5}$	FLOWER
Initialisation	released pi05_base	Florence-2-large with the official $360$ k-step pretrained weights
Global batch	$8$	$32$
Steps	$30$ k	$30$ k ( $30$ epochs of $1000$ )
Reported checkpoint	ours $10$ k, vanilla $30$ k	ours epoch $10$ ( $\approx\!10$ k), vanilla epoch $30$
Learning rate	$5\times 10^{-5}\!\rightarrow\!5\times 10^{-6}$ cosine, $1500$ warmup	$2\times 10^{-5}$ , the host’s three-stage schedule, weight decay $0.05$
Optimiser	AdamW, gradient clipping $1.0$	AdamW, betas $(0.9,0.99)$

Appendix

Table 6: Training settings for the two transfer hosts. Unlisted settings follow the published defaults.

Method	Spatial	Object	Goal	Long	Avg.	$H_{\mathrm{exec}}$	Act./s	VRAM (GB)
S-WAM (LaWAM)	98.6	100.0	97.0	95.0	97.7	50	292.7	5.2
MiniCPM-RobotManip ( OpenBMB, 2026 )	–	–	–	–	97.5	30	166.8	3.7
JEPA-WAM ( Lin et al., 2026b )	95.6	99.4	97.2	94.6	96.7	20	150.3	4.3
Evo-Depth ( Lin et al., 2026a )	95.6	99.2	95.6	91.3	95.4	50	145.5	3.0
$\pi_{0.5}$ ( Intelligence et al., 2025 ) †	98.8	98.2	98.0	92.4	96.9	5	132.1	9.5
OpenVLA-OFT ( Kim et al., 2025 )	97.6	98.4	97.9	94.5	97.1	8	114.7	16.1

Appendix

Table 7: LIBERO success rate and measured throughput across methods.

Figure 9: Per-suite breakdown of Fig. 6 .

Method	Spatial	Object	Goal	Long	Avg.
TurboVLA	97.4	99.4	96.2	93.2	96.5
VLA-Adapter	96.6	99.8	95.8	84.0	94.0
Evo-1	92.8	98.2	92.4	86.4	92.5
Vanilla, $H_{\mathrm{exec}}=10$	96.2	98.4	95.6	90.6	95.2
Vanilla, full chunk ( $H=50$ )	87.2	80.6	88.0	73.8	82.4
Ours, full chunk ( $H=50$ )	98.6	100.0	97.0	95.0	97.7

Appendix

Table 8: Detailed results on LIBERO. Avg. is the unweighted mean over the four suites.

Method	Background	Robot	Camera	Language	Noise	Layout	Light	Avg.
TurboVLA	77.1	30.3	76.0	72.2	71.2	60.0	81.5	66.9
VLA-Adapter	89.0	38.4	89.3	66.6	91.9	73.1	87.5	76.5
Evo-1	92.6	37.7	87.2	61.1	89.7	60.9	91.7	74.4
Vanilla, $H_{\mathrm{exec}}{=}50$	83.0	49.7	77.1	53.9	80.0	64.8	85.7	70.6
Vanilla, $H_{\mathrm{exec}}{=}10$	95.2	73.4	91.5	77.7	94.3	82.1	96.0	87.2
S-WAM (ours), $H_{\mathrm{exec}}{=}50$	96.6	76.0	89.5	85.2	90.4	79.3	98.6	87.9

Appendix

Table 9: Per-category LIBERO-Plus results averaged over the four LIBERO suites. Avg. is the unweighted mean of the seven categories.

Method	Background	Robot	Camera	Language	Noise	Layout	Light	Avg.
TurboVLA	91.5	28.3	69.9	82.8	68.9	54.3	92.5	69.7
VLA-Adapter	98.8	50.0	95.7	75.4	98.9	93.0	98.3	87.2
Evo-1	92.2	36.9	87.8	68.5	91.5	63.6	89.7	75.7
Vanilla, $H_{\mathrm{exec}}{=}50$	88.0	49.4	83.0	53.3	80.1	78.2	94.2	75.2
Vanilla, $H_{\mathrm{exec}}{=}10$	98.4	73.1	95.7	81.5	97.2	93.0	98.3	91.0
S-WAM (ours), $H_{\mathrm{exec}}{=}50$	98.1	75.7	93.1	87.2	94.3	85.2	98.6	90.3

Appendix

Table 10: Per-category LIBERO-Plus results on LIBERO-Spatial . Avg. is the unweighted mean of the seven categories.

Method	Background	Robot	Camera	Language	Noise	Layout	Light	Avg.
TurboVLA	99.6	36.7	100.0	99.4	99.8	78.4	99.7	87.7
VLA-Adapter	96.0	27.1	97.2	84.5	96.9	75.7	95.3	81.8
Evo-1	96.8	27.1	95.2	77.7	93.4	72.2	98.7	80.1
Vanilla, $H_{\mathrm{exec}}{=}50$	84.3	35.7	78.3	58.5	84.1	65.0	89.2	70.7
Vanilla, $H_{\mathrm{exec}}{=}10$	99.6	72.1	98.5	81.1	98.6	91.1	99.7	91.5
S-WAM (ours), $H_{\mathrm{exec}}{=}50$	96.8	71.9	94.7	88.1	97.4	83.4	100.0	90.3

Appendix

Table 11: Per-category LIBERO-Plus results on LIBERO-Object . Avg. is the unweighted mean of the seven categories.

Method	Background	Robot	Camera	Language	Noise	Layout	Light	Avg.
TurboVLA	45.9	10.5	54.2	28.3	40.6	35.5	47.7	37.5
VLA-Adapter	92.5	42.5	91.2	53.7	93.4	59.1	82.1	73.5
Evo-1	91.5	39.6	85.7	44.9	85.8	50.4	92.1	70.0
Vanilla, $H_{\mathrm{exec}}{=}50$	87.5	61.9	81.9	48.3	83.9	58.8	82.1	72.1
Vanilla, $H_{\mathrm{exec}}{=}10$	94.0	78.7	90.0	67.8	92.9	63.1	90.7	82.4
S-WAM (ours), $H_{\mathrm{exec}}{=}50$	95.4	80.0	89.5	78.5	90.5	67.1	97.8	85.5

Appendix

Table 12: Per-category LIBERO-Plus results on LIBERO-Goal . Avg. is the unweighted mean of the seven categories.

Method	Background	Robot	Camera	Language	Noise	Layout	Light	Avg.
TurboVLA	71.3	45.5	79.7	78.3	75.5	71.8	86.1	72.6
VLA-Adapter	68.5	33.8	73.0	53.0	78.4	64.7	74.5	63.7
Evo-1	90.0	47.3	80.1	53.3	88.4	57.4	86.1	71.8
Vanilla, $H_{\mathrm{exec}}{=}50$	72.3	51.7	65.4	55.4	71.7	57.1	77.4	64.4
Vanilla, $H_{\mathrm{exec}}{=}10$	88.9	69.5	81.6	80.4	88.6	81.4	95.3	83.7
S-WAM (ours), $H_{\mathrm{exec}}{=}50$	96.2	76.3	80.7	86.9	79.5	81.7	97.8	85.6

Appendix

Table 13: Per-category LIBERO-Plus results on LIBERO-Long . Avg. is the unweighted mean of the seven categories.

Method	adjust bottle	beat block	click alarm	click bell	grab roller	move can	move card	press stapler	rotate QR	Avg.
TurboVLA	0	0	8	0	0	0	0	2	0	1.11
VLA-Adapter	10	0	2	0	24	6	0	6	0	5.33
Evo-1	0	0	6	0	0	0	0	2	0	0.89
Vanilla, $H_{\mathrm{exec}}=10$	0	0	0	0	12	0	0	8	0	2.22
Vanilla, full chunk ( $H=75$ )	66	6	4	0	30	14	6	14	6	16.22
Ours, full chunk ( $H=75$ )	60	24	12	2	38	2	12	20	4	19.33

Appendix

Table 14: Task-level success rates (%) on DOMINO. Avg. is the macro-average over the nine tasks.

Method	Act/s (p50)	TTFA (ms)	VRAM (GB)
TurboVLA	340.3	35.3	0.5
VLA-Adapter	91.3	87.7	3.6
Evo-1	41.9	334.1	2.4
Vanilla, $H_{\mathrm{exec}}=10$	80.9	123.6	5.2
Vanilla, $H_{\mathrm{exec}}=50$	402.5	124.2	5.2
Ours, $H_{\mathrm{exec}}=50$	292.7	73.3	5.2

Appendix

Table 15: Measured inference efficiency across methods.

Task	Demos	Frames	Length	Span	Prompt
Stack bowl (static)	30	$27{,}944$	19.1 s	90%	stack the red bowl on the blue bowl
Stack bowl (dynamic)	30	$23{,}073$	14.6 s	69%	stack the red bowl on the blue bowl
Hang cup (static)	30	$25{,}958$	17.0 s	52%	hang the yellow cup on the cup rack
Hang cup (dynamic)	30	$32{,}247$	21.2 s	49%	hang the yellow cup on the cup rack
Place corn in bowl (static)	30	$30{,}969$	18.1 s	99%	place the corn in the bowl
Place corn in bowl (dynamic)	30	$24{,}838$	16.1 s	78%	place the corn in the bowl

Appendix

Table 16: The nine real-robot datasets. Length denotes the median episode duration. Span denotes the range over which manipulated objects are re-placed between demonstrations, measured as a percentage of image width. All data is recorded at

50

Hz.

Task	Counted as a success when
Stack bowl	the red bowl rests inside the blue bowl after release without tipping or falling outside
Hang cup	the cup remains on the rack arm after release
Place corn in bowl	the yellow corn is placed inside the bowl and remains there; grasping another object is a failure
Pour water	the water is poured into the blue cup without dropping the red cup or pouring outside the target
Put lid on the cup	the lid rests on the cup rim and covers the opening without falling onto the table
Place bowl in drawer	the drawer is opened and the bowl is placed inside; both stages are required

Appendix

Table 17: Success criteria for the real-robot evaluation. The static and dynamic versions of a task share the same criterion.

Panel of Figure 10 — Figure 10: Stack bowl, static. The robot picks up the red bowl and places it inside the blue bowl.

Panel of Figure 14 — Figure 14: Place corn in bowl, static. The robot picks up the yellow corn and places it in the bowl in the presence of distractor objects.

Task	Setting	Vanilla, $H_{\mathrm{exec}}{=}20$	Vanilla, $H_{\mathrm{exec}}{=}50$	S-WAM
Stack bowl	Static	85.0	72.5	90.0
Hang cup	Static	57.5	52.5	70.0
Place corn in bowl	Static	77.5	60.0	75.0
Put lid on the cup	Static	75.0	67.5	80.0
Pour water	Static	62.5	50.0	67.5
Place bowl in drawer	Static	47.5	40.0	52.5

Appendix

Table 18: Real-robot success rate (%) per task, each cell over

40

trials. S-WAM and vanilla at

H_{\mathrm{exec}}{=}50

are compute-matched; vanilla at

H_{\mathrm{exec}}{=}20

replans

2.5\times

as often.

Explore similar work

Streaming-WAM: Action-Conditioned World-Action Model for Asynchronous Robot Manipulation
Sep 24, 2026Xuyao Huang, Yixuan Wang, Zengyao Ye +4Action Chunks Robotic Manipulation
SJTU · SDU · ECUST +1
Light-WAM: Efficient World Action Models with State-Fusion Action Decoding
Jun 6, 2026Ziang Li, Dongzhou Cheng, Yibin Wang +5Efficient World-Action Model Action Prediction
Wuhan University · Shanghai Innovation Institute · Southeast University +2
SplineWAM: Adaptive Action Horizons for World Action Models via B-Spline Representations
Sep 30, 2026Jun Guo, Xiaoshen Han, Qiwei Li +7Action Chunks World Models
Tsinghua University · Xiaomi Robotics · Shanghai Jiao Tong University +2

Sep 24, 2026, cs.ROJ/K move · Enter open · S save

Streaming-WAM: Action-Conditioned World-Action Model for Asynchronous Robot Manipulation

Xuyao Huang, Yixuan Wang, Zengyao Ye, Boyuan Zhao+3

SJTU · SDU · ECUST+1

Action Chunks Robotic Manipulation Robot Systems

Open paper AskPDF arXiv

World action models (WAMs) that use future visual prediction at inference time incur substantial generation costs. Asynchronous execution reduces waiting by overlapping inference with robot motion, but visual predictions used for subsequent action generation must anticipate the effects of actions already scheduled for execution during inference. We introduce Streaming-WAM, which couples action-conditioned world modeling with asynchronous robot control to account for committed actions in future visual prediction. At each streaming update, the model conditions future visual prediction on the latest observation and the committed actions, which form the fixed prefix of the next action chunk. The resulting action-conditioned visual features guide generation of the remaining actions within the same joint update, so the continuation is informed by the scene changes expected during execution of the fixed prefix. On LIBERO, Streaming-WAM achieves an average success rate of 98.35% and reduces mean episode time by a factor of 2.93 relative to Fast-WAM. On the real-world Stamp Paper task, mean episode time falls from 90 s with synchronous Joint-WAM to 38 s with Streaming-WAM. These results show that Streaming-WAM supports efficient asynchronous control while maintaining high task success rates.

Abstract

Explore similar work

Streaming-WAM: Action-Conditioned World-Action Model for Asynchronous Robot Manipulation

Light-WAM: Efficient World Action Models with State-Fusion Action Decoding

SplineWAM: Adaptive Action Horizons for World Action Models via B-Spline Representations