Learning from human demonstration (LfD), among many speedup techniques for reinforcement learning (RL), has seen many successful applications. We consider one LfD technique called human–agent transfer (HAT), where a model of the human demonstrator’s decision function is induced via supervised learning and used as an initial bias for RL. Some recent work in LfD has investigated learning from observations only, that is, when only the demonstrator’s states (and not its actions) are available to the learner. Since the demonstrator’s actions are treated as labels for HAT, supervised learning becomes untenable in their absence. We adapt the idea of learning an inverse dynamics model from the data acquired by the learner’s interactions with the environment and deploy it to fill in the missing actions of the demonstrator. The resulting version of HAT—called state-only HAT (SoHAT)—is experimentally shown to preserve some advantages of HAT in benchmark domains with both discrete and continuous actions. This paper also establishes principled modifications of an existing baseline algorithm—called A3C—to create its HAT and SoHAT variants that are used in our experiments.

Through diversity of composition, sequence, and interfacial structure, hybrid materials greatly expand the palette of materials available to access novel functionality. The NSF Division of Materials Research recently supported a workshop (October 17–18, 2019) aiming to (1) identify fundamental questions and potential solutions common to multiple disciplines within the hybrid materials community; (2) initiate interfield collaborations between hybrid materials researchers; and (3) raise awareness in the wider community about experimental toolsets, simulation capabilities, and shared facilities that can accelerate this research. This article reports on the outcomes of the workshop as a basis for cross-community discussion. The interdisciplinary challenges and opportunities are presented, and followed with a discussion of current areas of progress in subdisciplines including hybrid synthesis, functional surfaces, and functional interfaces.

Military attachés and wartime observers have received surprisingly little attention in international relations. Why do states exchange attachés, permitting uniformed foreigners to gather intelligence on their territory and during their wars? To explain, we adopt a broadly practice-theoretic approach, focusing on the individuals who developed the role by living it, showing how they both innovated a distinct military practice and established institutional legitimacy for attachés. We address an early historical case in which the practice proliferated: the Russo-Japanese War, throughout which observers represented multiple European states, on both sides of the conflict. Sometimes termed the first modern war, the conflict saw Japan's entry into the Eurocentric great power system. In this context, embedded attachés had a dual effect. On the one hand, a professional attaché community established itself: we show how local innovation by embedded officers, in the context of this structurally destabilising event, permitted the creation of a new institutional role that might otherwise have been impossible. On the other, the Japanese made use of the attachés as witnesses for Western governments, observing their performance of great power-hood, as they defeated Russia. The argument has implications for understanding both the military attaché system and communities of practice as such.

A volume-area scaling relation is commonly used to estimate glacier volume or its future changes on a global scale. The presence of an insulating supraglacial debris cover alters the mass-balance profile of a glacier, potentially modifying the scaling relation. Here, the nature of scaling relations for extensively debris-covered glaciers is investigated. Theoretical arguments suggest that the volume-area scaling exponent for these glaciers is ~7% smaller than that for clean glaciers. This is consistent with the results from flowline-model simulations of idealised glaciers, and the available data from the Himalaya. The best-fit scale factor for debris-covered Himalayan glaciers is ~60% larger compared to that for the clean ones, implying a significantly larger stored ice volume in a debris-covered glacier compared to a clean one having the same area. These results may help improve scaling-based estimates of glacier volume and future glacier changes in regions where debris-covered glaciers are abundant.

Frailty is increasingly used in clinical settings to describe a physiological state resulting from a combination of age-related co-morbidities. Frailty also has a strong ‘lay’ meaning that conjures a particular way of being. Recent studies have reported how frail older people perceive the term frailty, showing that frailty is often an unwanted and resisted label. While there are many scores and measures that clinicians can use to determine frailty, little has been published regarding how health-care professionals use and make sense of the term. This paper reports the findings of a qualitative study that explored how health professionals perceive frailty. Forty situated interviews were conducted with health-care professionals working in an emergency department in the English Midlands. The interview talk was analysed using discourse analysis. The findings show that the health professionals negotiate an ‘ideological dilemma’ – a tension between contradictory sets of meanings and consequences for action – based on their ‘lay’ and clinical experience of the term frailty. It is concluded that this dilemma could have a negative impact on the assessment of frailty depending on the system of assessment used.