Boundary Graph
An improved method for constructing attribution graphs that adds first-order variations to pre-activation values and re-sorts them, so as to capture the dynamics of features entering and leaving the active set and to improve prediction accuracy.
Timeline (last 90 days)
The researchers proposed the "margin graph": when constructing the graph, a first-order change term is added to each pre-activation value and then re-sorted; this method has the same construction cost as the original graph but significantly lower error.
In a write operation that scales feature encodings by three times, the graph built from previously active features has a median relative error of at least 0.247, while the maximum error of the margin graph is only 0.033.
All Facts (2)
The researchers proposed the "margin graph": when constructing the graph, a first-order change term is added to each pre-activation value and then re-sorted; this method has the same construction cost as the original graph but significantly lower error.
50%UnverifiedIn a write operation that scales feature encodings by three times, the graph built from previously active features has a median relative error of at least 0.247, while the maximum error of the margin graph is only 0.033.
50%