Embodied control increasingly requires models to follow compositional language instructions while reasoning over dynamic visual states. However, current vision-language-action policies and world-action models often couple linguistic knowledge with visual computation in a shared backbone or conditioning pathway, leading to modality competition and making knowledge extension dependent on backbone updates. In this paper, we introduce Key-Gram, a conditional-memory framework that separates language-derived world knowledge from visual-state reasoning for embodied control. At its core is a memory module that decomposes an instruction into task-specific key-grams, retrieves static linguistic priors through deterministic hashed lookup, and injects the retrieved entries into selected hidden layers through context-aware gating and lightweight convolutional fusion. This design allows the backbone to devote its main capacity to visual reasoning and action inference, while reusable instruction knowledge is stored in an extensible external memory. The logical memory table can be conveniently partitioned during training and, due to its $O(1)$ lookup pattern, efficiently placed on host memory during inference. Across RoboTwin2.0, LIBERO/LIBERO-Plus, and real-world dual-arm manipulation, Key-Gram consistently improves both $π_{0}$ and $π_{0.5}$ backbones, with average relative gains of $29.5\%/9.9\%$ on RoboTwin2.0, $35.8\%/4.5\%$ on LIBERO-Plus transfer without target-domain fine-tuning, and $15.4\%/8.1\%$ on real-world long-horizon tasks. These results demonstrate that externalized linguistic memory provides an effective and extensible mechanism for improving compositional grounding, transfer, and real-world manipulation.

翻译：具身控制日益要求模型在推理动态视觉状态的同时遵循组合语言指令。然而，当前的视觉-语言-动作策略和世界-动作模型往往将语言知识与视觉计算耦合在共享主干或条件路径中，导致模态竞争，并使知识扩展依赖于主干更新。本文提出Key-Gram，一种条件记忆框架，将源自语言的世界知识与具身控制的视觉状态推理相分离。其核心是一个记忆模块，该模块将指令分解为任务特定的关键语法单元，通过确定性的哈希查找检索静态语言先验，并通过上下文感知门控和轻量级卷积融合将检索到的条目注入选定的隐藏层。这种设计使主干能够将其主要容量用于视觉推理和动作推理，而可重用的指令知识则存储于可扩展的外部记忆中。逻辑记忆表可在训练期间便捷地分区，且由于其O(1)查找模式，在推理时可高效地放置于主机内存。在RoboTwin2.0、LIBERO/LIBERO-Plus以及真实世界双臂操作任务中，Key-Gram持续改进了π₀和π₀.₅主干，在RoboTwin2.0上平均相对增益分别为29.5%/9.9%，在无需目标域微调的LIBERO-Plus迁移任务上为35.8%/4.5%，在真实世界长时域任务上为15.4%/8.1%。这些结果表明，外部化的语言记忆为提升组合接地、迁移和真实世界操作提供了一种有效且可扩展的机制。