空间记忆需存储几何信息，以应对遮挡场景下的记忆检索。
What Spatial Memory Must Store: Occlusion as the Test for Language-Agent Memory
- 用几何权重替代线性混合，显著提升空间查询下的记忆召回率。
- 分离回忆与可见性：通过射线-体素分析，准确识别被遮挡目标。
- 实测验证了遮挡判断必须依赖几何数据，适合智能体环境建模研究者。
语言智能体的‘记忆宫殿’系统将每条记忆锚定在世界坐标上，基于几何信息能补充文本无法表达的内容这一直觉。我们使该直觉可测试，并报告三项结果：第一，默认将空间邻近性线性融合进回忆权重的方法无益甚至有害；在预注册的召回实验中，该方法在冻结测试中表现平庸（平均Delta-Hit@5 -0.0375，Wilcoxon p=0.306），而几何主导的加权方式则显著胜出（+0.3208，p<10^-15）；第二，回忆应无视遮挡（正确记住墙后房间），而可见性是基于存储几何的感知谓词，实时系统无需计算；使用一条重定向自代理视线射线的射线-体素数字微分分析器（DDA），文本与实时视野锥均对849个墙后目标得分为0.000，而锥+DDA达到0.982（精确McNemar p<10^-6）；坐标召回能区分近似重复位置，优于余弦相似度（1.000 vs 0.533，n=150）；第三，该可见性谓词在提交git的预注册实验（SPMEM-OCC-LIVE-v1）中被实时验证：八个脚本世界，自动评分，96个墙后目标，误显率从1.000降至0.000，合并精确McNemar p=2.5x10^-29，该运行还发现并修复了一个真实锚点缺陷。我们承认遮挡需要几何信息近乎同义反复；贡献在于测量与隔离，明确空间记忆必须存储什么，而非如何读取。这些初步实验支撑一个冻结确认研究（SPMEM-ZERO-REAL-PREREG-v1）；包含盲评的人类撰写多世界研究尚待未来完成。
原文摘要 · Abstract (English)
Language-agent "memory palace" systems anchor each memory to a world coordinate, on the intuition that geometry adds something text cannot. We make that intuition testable and report three results. First, the memory-palace default of folding spatial proximity into a linear blend beside recency and importance does not help and can hurt: in a pre-registered recall experiment the shipped blend fails its own frozen test (mean Delta-Hit@5 -0.0375, Wilcoxon p=0.306), sitting at a position-blind baseline, while a geometry-led weighting wins decisively (+0.3208, p<10^-15): geometry must lead recall when the query regime is spatial. Second, memory recall and visibility must be separated: recall is occlusion-blind by design (you correctly remember the next room behind a wall), while visibility is a perception predicate over stored geometry that the live system never computed. A one-line ray-versus-voxel digital differential analyzer (DDA), re-pointed from the gaze ray the agent already casts, supplies it: text and the live FoV cone both score 0.000 on 849 behind-wall targets while cone-plus-DDA reaches 0.982 (exact McNemar p<10^-6); coordinate recall separately resolves near-duplicate locations a cosine null cannot (1.000 vs 0.533, n=150). Third, the visibility predicate is confirmed live under a git-committed pre-registration (SPMEM-OCC-LIVE-v1: eight scripted worlds, automated oracle scoring, 96 behind-wall targets, false-visible 1.000->0.000, pooled exact McNemar p=2.5x10^-29), a run that surfaced and fixed a real relay anchor defect. We concede that occlusion-needs-geometry is near-tautological; the contribution is the measurement and isolation, separating what spatial memory must store from how it is read. These pilots power a frozen confirmatory study (SPMEM-ZERO-REAL-PREREG-v1); the full human-authored multi-world study with blind raters remains future work.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。