Question answering (QA) models for reading comprehension tend to exploit spurious correlations in training sets and thus learn shortcut solutions rather than the intended by QA datasets. that have learned can achieve human-level performance examples where shortcuts are valid, but these same behaviors degrade generalization potential on anti-shortcut invalid. Various methods been proposed mitiga...

Place cells are hippocampal neurons whose discharge is strongly related to a rat's location in its environment. The existence of place cells has led to the proposal that they are part of an integrated neural system dedicated to spatial navigation. To further understand the relationships between place cell firing and spatial problem solving, we examined the discharge of CA1 and CA3 place cells a...

The physical basis of shortcut edges in a Protein Residue Network (PRN) and their applicability to a simplified protein folding pathway problem are investigated. The results show promise and suggest the dominant influence of native-state topology even for structurally homologous proteins with different preferred folding pathways. Shortcut edges are enriched with hydrogen bonds, and there is a s...

In this paper, a novel architecture for a deep recurrent neural network, residual LSTM is introduced. A plain LSTM has an internal memory cell that can learn long term dependencies of sequential data. It also provides a temporal shortcut path to avoid vanishing or exploding gradients in the temporal domain. The residual LSTM provides an additional spatial shortcut path from lower layers for eff...

We study a graph-augmentation problem arising from a technique applied in recent approaches for route planning. Many such methods enhance the graph by inserting shortcuts, i.e., additional edges (u,v) such that the length of (u,v) is the distance from u to v. Given a weighted, directed graph G and a number c ∈ Z>0, the shortcut problem asks how to insert c shortcuts into G such that the expecte...

| To eeciently transfer large amounts of diverse traac over high-speed links, modern integrated networks require more eecient packet-switching techniques that can capitalize on recent advances in switch hardware. Several promising approaches attempt to improve performance by creating dedicated \shortcut" connections for long-lived traac ows, at the expense of the network overhead for establishi...

Partial expected value of perfect information (EVPI) calculations can quantify the value of learning about particular subsets of uncertain parameters in decision models. Published case studies have used different computational approaches. This article examines the computation of partial EVPI estimates via Monte Carlo sampling algorithms. The mathematical definition shows 2 nested expectations, ...