RealtimeWAM: How Fast Can I Run My World Action Model?
RealtimeWAM: How Fast Can I Run My World Action Model?
This paper presents RealtimeWAM, a general, training-free framework for low-latency inference in World Action Models (WAMs). It addresses the high inference latency of WAMs by coordinating parallel execution with adaptive computation, overlapping observation processing with prediction, and selectively reusing cached features. Evaluated on FastWAM and OpenWAM across RoboTwin, LIBERO, and LIBERO-Plus, RealtimeWAM achieves speedups of 8.90× and 10.67× with average success rates of 82.75% and 87.41%, respectively.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
mean inference latencies
24.09
mean inference latencies
63.09
average speedups
8.9
average speedups
10.67
average success rates
82.75
average success rates
87.41
improves average success rates
17.2
View original evidence
RealtimeWAM improves average success rates over native inference by 17.2 and 37.2 percentage points on FastWAM and OpenWAM
Open source S5improves average success rates
37.2
View original evidence
RealtimeWAM improves average success rates over native inference by 17.2 and 37.2 percentage points on FastWAM and OpenWAM
Open source S5
What remains unknown
Not established in the collected evidence: Environment, Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
lus. On an RTX 4090, measured mean inference latencies are 24.09 and 63.09 ms, corresponding to average speedups of 8.90$\times$ and 10.67$\times$. Average success rates are 82.75% and 87.41%, respectively, within 0.02 and 0.53 percentage points of native inference. Across five real-world tasks, RealtimeWAM improves average success rates over native inference by 17.2 and 37.2 percentage points on FastWAM and OpenWAM,
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org