发表机构
RAWS Labs（RAWS实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
TinyCast是仅14.65万参数的无注意力零样本预测模型，通过计算周期建模，在多基准上参数远小于同类模型，可在嵌入式设备端到端预测。
AI 中文摘要
我们提出了TinyCast，一种无注意力机制的零样本预测模型，其仅使用146505个参数即可输出预测分布，前提是在该规模下，上下文的周期结构值得被计算而非学习。一个零参数的频谱检测器提供主导周期，上下文根据其相位进行折叠，然后通过一个膨胀卷积编码器和一个块自回归分位数解码器对剩余部分进行建模。它比GIFT-Eval平台上所有可确定参数数量的零样本预测模型都要小。在概率准确性方面，它定义了规模-准确性前沿。在所有宣称无测试数据泄漏的零样本模型中，它是唯一参数低于140万且能输出预测分布的模型，且所有得分高于它的模型至少需要该预算。在Chronos-ZS和fev-bench数据集上，所有排名在它之前的神经模型的参数至少是它的28倍。由于其混合路径仅由卷积和矩阵乘法构成，它可导出为静态INT8格式，并能在嵌入式设备上进行端到端预测，无需针对每个信号进行拟合。
英文摘要
We introduce TinyCast, an attention-free zero-shot forecaster that emits a predictive distribution from 146,505 parameters, on the premise that at this size the periodic structure of a context is worth computing rather than learning. A zero-parameter spectral detector supplies the dominant periods, the context is folded on their phase, and a dilated convolutional encoder and a block-autoregressive quantile decoder model the rest. It is smaller than every zero-shot entry on the GIFT-Eval board whose parameter count can be established. On probabilistic accuracy it defines the size-accuracy frontier. Among zero-shot entries declaring no test-data leakage it is the only one below 1.4M parameters that emits a predictive distribution, and every entry scoring better carries at least that budget. On Chronos-ZS and fev-bench every neural model ahead of it carries at least 28 times its parameters. Because the mixing path is convolutions and matrix multiplications only, it exports to static INT8 and forecasts end to end on an embedded device without per-signal fitting.
Comments38 pages, 6 figures, 17 tables. Code and weights: https://github.com/raws-labs/tinycast
