---
title: "Leap-VLA: High-Rate Robot Control from a Low-Rate VLM Backbone"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.01011v1"
permalink: "https://feed.astralor.com/items/itm_3b46365328af9370637d2f58"
published: "2026-10-31T14:53:16.000Z"
score: 59
channel: "tech"
tags: ["推理效率","RAG","多模态","Leap-VLA","High-Rate Robot Control","Low-Rate VLM Backbone","Vision-language-action","VLA"]
---

# Leap-VLA: High-Rate Robot Control from a Low-Rate VLM Backbone

## At a glance

Vision-language-action (VLA) models re-encode the full observation through a multi-billion-parameter vision-language model (VLM) at every control step, even though scene semantics evolve at roughly 1 Hz while reactive c…

## Key points

- Vision-language-action (VLA) models re-encode the full observation through a multi-billion-parameter vision-language model (VLM) at every control step, even though scene semantics evolve at roughly 1 Hz while reactive c…
- We present a VLA that decouples the two: a VLM runs as a low-frequency semantic anchor whose cached context is reused across control steps, while a flow-matching action expert operates at high frequency, consuming the c…
- The measured age $τ$ of the cache conditions the expert through adaptive layer normalization.

## Full text

Vision-language-action (VLA) models re-encode the full observation through a multi-billion-parameter vision-language model (VLM) at every control step, even though scene semantics evolve at roughly 1 Hz while reactive control requires 20-50 Hz. We present a VLA that decouples the two: a VLM runs as a low-frequency semantic anchor whose cached context is reused across control steps, while a flow-matching action expert operates at high frequency, consuming the cached semantics with a fresh, low-latency visual stream from a compact encoder pretrained on robot video with masked reconstruction, optical flow, and depth. The measured age $τ$ of the cache conditions the expert through adaptive layer normalization. Reusing the cache makes the policy a function of its age, and training covers that function explicitly: cache ages sampled from real episode history, plus an age-ramped pixel corruption that erases the stale snapshot the way deployed staleness corrupts content. On the asynchronous deployment stack, the recipe closes most of the staleness gap: on LIBERO-Spatial, 61.4% to 95.5% at a 2 s semantic refresh (95.9% at 1 s, within 1.6 points of the frequency-matched anchor); averaged over the four LIBERO suites, the deployed policy holds 95.2% at a 1 s refresh and 94.5% at 2 s against a 96.3% matched anchor. A forward-only chunk-boundary correction (soft-nudge), an autograd-free reformulation of real-time-chunking guidance, carries per-control-step re-planning without loss (97.1%) at zero added latency. The same lens exposes a second defect: under an instruction-scene mismatch the policy keeps manipulating the scene, and a chain-of-thought verdict gate trained on zero-cost negative episodes freezes execution exactly on mismatched instructions without hurting feasible success.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_3b46365328af9370637d2f58