Search

The ability to derive useful information by asking clarifying questions (ACQ) is an important element of real life collaboration on reasoning tasks, such as question answering (QA). Existing natural language ACQ challenges, however, evaluate generations based on word overlap rather than the value of the information itself. Word overlap is often an inappropriate metric for question generation since many different questions could be useful in a given situation, and a single question can be phrased many different ways. Instead, we propose evaluating questions pragmatically based on the value of the information they retrieve. Here we present a definition and framework for natural language pragmatic asking of clarifying questions (PACQ), the problem of generating questions that result in answers useful for a reasoning task. We also present fact-level masking (FLM), a procedure for converting natural language datasets into self-supervised PACQ datasets by omitting particular critical facts. Finally, we generate a PACQ dataset from the HotpotQA dataset using FLM and evaluate several zero-shot language models on it. Our experiments show that current zero-shot models struggle to ask questions that retrieve useful information, as compared to human annotators. These results demonstrate an opportunity to use FLM datasets and the PACQ framework to objectively evaluate and improve question generation and other language models.

We present a novel device that can offer two extremes of elastic wave propagation --- nearly complete transmission and strong attenuation under impulse excitation. The mechanism of this highly tunable device relies on intermixing effects of dispersion and nonlinearity. The device consists of identical cylinders arranged in a chain, which interact with each other as per nonlinear Hertz's contact law. For a `dimer' configuration, i.e., two different contact angles alternating in the chain, we analytically, numerically, and experimentally show that impulse excitation can either propagate as a localized wave, or it can travel as a highly dispersive wave. Remarkably, these extremes can be achieved in this periodic arrangement simply by \textit{in-situ} control of contact angles between cylinders. We close the discussion by highlighting the key characteristics of the mechanisms that facilitate strong attenuation of incident impulse. These include low frequency to high frequency (LF-HF) scattering, and turbulence-like cascading in a periodic system. We thus envision that these adaptive, cylinder-based nonlinear phononic crystals, in conjunction with conventional impact mitigation mechanisms, could be used to design highly tunable and efficient impact manipulation devices.