AtlasVLA：視覺-語言-行動模型的持久世界-自我狀態建模

摘要

雖然視覺-語言-動作（VLA）模型推動了具身人工智能的發展，但其根本上的反應式範式嚴重限制了在部分可觀測及長時程任務中的表現。當僅限於單一腕戴式相機時，此類模型不可避免地面臨兩大問題：物體離開視野時產生的感知遺忘，以及多步驟執行過程中的時序任務進度遺忘。為克服這些瓶頸，我們提出AtlasVLA——一個新穎框架，透過持久世界-自我狀態，將直接反應式操控轉變為主動推理。AtlasVLA具備雙記憶架構：4D持久世界狀態記憶，將短暫的2D觀測提升為全域更新、以體素哈希儲存的空間狀態，以解決視覺盲區；以及自我工作狀態記憶，用於追蹤歷史自我狀態與任務進度。透過將擴散變換器（DiT）以聯合世界-自我狀態為條件，AtlasVLA實現了穩健的空間推理。在LIBERO、RLBench及真實世界基準上的廣泛評估表明，AtlasVLA僅使用腕戴式相機即達到了最先進的性能。尤為值得注意的是，它大幅超越了多視角基線方法，在LIBERO-Long上取得了9.4%的絕對成功率提升，在真實世界長時程任務中則提升了17.5%。

English

While Vision-Language-Action (VLA) models have advanced embodied AI, their fundamentally reactive paradigm severely limits performance in partially observable and long-horizon tasks. When restricted to a single wrist-mounted camera, they inevitably suffer from perception forgetting as objects exit the field of view, and temporal task-progress forgetting} during multi-step execution. To overcome these bottlenecks, we propose AtlasVLA, a novel framework that transitions from direct reactive manipulation to proactive reasoning through a persistent world-ego state. AtlasVLA features a dual-memory architecture: a 4D Persistent World State Memory that lifts transient 2D observations into a globally updated, voxel-hashed spatial state to resolve visual blind spots, and an Ego-Working State Memory that tracks historical ego state and task progress. By conditioning a diffusion transformer (DiT) on this joint World-Ego state, AtlasVLA enables robust spatial reasoning. Extensive evaluations across LIBERO, RLBench, and real-world benchmarks demonstrate that AtlasVLA achieves state-of-the-art performance using solely a wrist camera. Remarkably, it decisively outperforms multi-view baselines, yielding absolute success rate improvements of 9.4% on LIBERO-Long and 17.5% in real-world long-horizon tasks.