DanLing NestedTensor: טנסורים מרוכבים נשלטים לחישובי למידה עמוק

תקציר מקורי באנגליתarXiv:2609.30379v1 Announce Type: new Abstract: Variable-size inputs are common in deep learning, but dense batching allocates a shared envelope and spends computation on padding. The cost multiplies across varying axes: an explicit pair state allocates $BN_{\max}^2$ positions instead of $\sum_i N_i^2$. Packing removes that waste, but composing packed operations still requires the logical axes and sample boundaries a flat buffer no longer exposes. We present DanLing NestedTensor, a PyTorch tensor abstraction that makes multi-ragged structure a property of the tensor itself. Packed values carry tensor-backed partitions and logical dimension order, so broadcasting creates ragged axes, feature transformations retain them, and reductions consume them. The same representation carries through au