发表机构
Tencent LIGHTSPEED（腾讯光速工作室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究通过动态FrozenLake测试平台，探究了VLM智能体的空间记忆过时问题，发现文本可解性不代表视觉接地、信任过时记忆存在安全隐患、审核无法完全消除差距，明确了相关核心挑战。
AI 中文摘要
借助记忆增强的VLM（视觉语言模型）智能体基于持久的空间知识开展行动，但随着环境变化，这些知识会悄然过时。我们探究当智能体必须协调自信的记忆断言与矛盾的观察结果时会发生什么，以及当前模型能否在冲突演变为与安全相关的错误前察觉它。我们采用动态FrozenLake测试平台，将过时检测任务与下游导航任务相结合，在文本和图像两种输入模式下，针对三个闭源模型和三个开源权重VLM开展实验（共1800次检测运行，以及四个LLM导航器在共享的50个随机种子规模下的12000次文本模式导航回合）。研究得出三项发现：其一，文本可解性不意味着视觉接地：能从文本中可靠标记过时条目的模型，在相同网格上的视觉F1分数跨度为0.887至0.067，最弱的模型仍会做出流畅、自信却忽略图像的决策；其二，不经审核就使用过时记忆是安全隐患：在我们的主要GPT-4o设置中，信任原始记忆的智能体死亡次数是完全不使用记忆的智能体的两倍多；其三，审核有帮助但无法消除差距：透明的读取时过滤器可消除文本模式下的大部分安全成本，但即使是最优的过时标签在当前网格规模下也未带来进一步显著增益，且当视觉审核不可靠时，过滤不会产生一致的益处。综上，这些结果将空间记忆过时界定为一种安全失效模式，并将记忆与观察冲突下可靠的视觉接地和动作选择，确立为记忆增强智能体的核心开放挑战。
英文摘要
Memory-augmented VLM agents act on persistent spatial knowledge, yet that knowledge silently goes stale as the environment changes. We ask what happens when an agent must reconcile a confident memory claim with a contradicting observation, and whether current models can catch the conflict before it becomes a safety-relevant mistake. Using a dynamic FrozenLake testbed, we pair a staleness-detection task with a downstream navigation task across three closed-source models and three open-weight VLMs under both text and image inputs (1,800 detection runs, and 12,000 text-mode navigation episodes over four LLM navigators at a shared 50-seed scale). Three findings emerge. First, text solvability does not imply visual grounding: models that flag stale entries reliably from text nonetheless span vision F1 from 0.887 down to 0.067 on the identical grids, and the weakest keeps making fluent, confident decisions that ignore the image. Second, consuming stale memory without an audit is a safety liability: in our primary GPT-4o setting, an agent that trusts raw memory dies more than twice as often as the same agent given no memory at all. Third, auditing helps but does not close the gap: a transparent read-time filter removes much of the safety cost in text mode, yet even oracle stale labels bring no further significant gain on the current grid size, and when visual auditing is unreliable, filtering yields no consistent benefit. Together these results frame spatial-memory staleness as a safety failure mode and isolate reliable visual grounding and action selection under memory--observation conflict as the central open challenges for memory-augmented agents.
