发表机构
Indiana University Indianapolis; Purdue University（印第安纳大学印第安纳波利斯分校; 普渡大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文研究事件流时间量化中硬分箱的不连续性，提出线性双箱插值编码器，其L1灵敏度最优且唯一，实验表明可降低47-72%表示漂移并保持准确率。
AI 中文摘要
事件流水线在学习前通常将异步时间戳离散化。此步骤看似无害，但其稳定性直接取决于时间分辨率。我们在表示层面研究这种依赖性。首先，我们证明硬时间分箱是不连续的：边界附近任意微小的时间戳偏移都可能导致单位事件质量在分箱之间移动。然后，我们定义了一类非负、质量守恒、分辨率保真的连续编码器，并证明该类中的每个编码器的全局L1灵敏度至少为2/Δ，其中Δ表示分箱宽度。线性双箱插值达到了这一极限。局部支持和一阶矩保持也使其具有唯一性。在SHD、N-MNIST和DVS128 Gesture上的实验支持了上述分析。在均匀时间戳预算下，线性插值将平均表示漂移降低了47-72%，同时保持干净准确率几乎不变。在DVS Gesture上，在所有测试预算和三个随机种子下，它产生了零预测翻转。在SHD上，测得的漂移遵循1/Δ，R²=0.992。
英文摘要
Event pipelines often discretize asynchronous timestamps before learning. This step looks harmless, but its stability depends directly on temporal resolution. We study this dependence at the representation level. We first show that hard temporal binning is discontinuous: an arbitrarily small timestamp shift near a boundary can move unit event mass between bins. We then define a class of nonnegative, mass-preserving, resolution-faithful continuous encoders and prove that every encoder in this class has global L1 sensitivity at least 2/Delta, where Delta denotes bin width. Linear two-bin interpolation attains this limit. Local support and first-moment preservation also make it unique. Experiments on SHD, N-MNIST, and DVS128 Gesture support the analysis. Across uniform timestamp budgets, linear interpolation lowers mean representation drift by 47-72% while keeping clean accuracy nearly unchanged. On DVS Gesture, it produces zero prediction flips across all tested budgets and three seeds. On SHD, measured drift follows 1/Delta with R^2 = 0.992.
