Papers › Understanding Asynchronous Inference Methods for Vision-Language-Action Models
Understanding Asynchronous Inference Methods for Vision-Language-Action Models
Ayoub Agouzoul
Title, abstract, authors and date from arXiv's metadata (CC0); this paper is not in the Papers with Code archive (frozen 2025-07-28).
Vision-Language-Action (VLA) models offer a promising path to generalist robot control, but their inference latency causes observation staleness when generated actions are executed asynchronously. Several methods have been proposed concurrently to mitigate this problem: inference-time inpainting (IT-RTC), training-time delay simulation (TT-RTC), future-state-aware conditioning (VLASH), and lightweight residual correction (A2C2). Each takes a fundamentally different approach, but they have so far been evaluated independently with different codebases, base policies, and protocols. We present a systematic comparison of these four methods under controlled conditions. We develop two unified codebases that integrate all methods with harmonized library and dataset versions, and we benchmark them on the Kinetix suite with MLPMixer policies and on the LIBERO manipulation benchmark with SmolVLA, sweeping inference delays up to d=20 control steps. A2C2's per-step residual correction is the most effective method on Kinetix, holding above 90% solve rate up to d=8, and also leads on LIBERO from d=4 onwards. IT-RTC is competitive at low delays but degrades sharply under long chunks (H=30) and high delays. TT-RTC is the most robust training-based method: stable across dₘₐₓ choices, generalizes beyond its training delay distribution, and adds zero inference overhead. VLASH exhibits a clear low-delay vs. high-delay trade-off governed by the fine-tuning delay range [0,dₘₐₓ]. Code is available at https://github.com/TheAyos/async-vla-inference
In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.
Code
Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.
Code Syntology ran Syntology
Syntology holds the repository link but has not harvested or run code from it.
Results from the paper
The Papers with Code archive ends with its 2025-07-28 snapshot. This paper's arXiv identifier, 2605.08168, was issued in May 2026, after that date, so the archive has no leaderboard rows for it.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections