Complexity of symbolic representation in working memory of Transformer correlates with the complexity of a task

Even though Transformers are extensively used for Natural Language Processing tasks, especially for machine translation, they lack an explicit memory to store key concepts of processed texts. This paper explores the properties of the content of symbolic working memory added to the Transformer model decoder. Such working memory enhances the quality of model predictions in machine translation task and works as a neural-symbolic representation of information that is important for the model to make correct translations. The study of memory content revealed that translated text keywords are stored in the working memory, pointing to the relevance of memory content to the processed text. Also, the diversity of tokens and parts of speech stored in memory correlates with the complexity of the corpora for machine translation task.     

Composite Deep Belief Network approach for enhanced Antepartum foetal electrocardiogram signal

Antepartum Foetal surveillance is the most vital epoch of investigation during the pregnancy period. This surveillance would provide an opening to plan and manage the Foetus during Intrapartum and Antepartum stages of pregnancy. Moreover, it will help to identify high risk Foetuses during pregnancies which are complicated by maternal health conditions like diabetes mellitus, intrauterine growth restriction, etc. The foetal electrocardiogram (fECG) signal can be detected in the course of pregnancy from the Antepartum stage. Generally, fECG signal analysis is not carried out for Foetal surveillance. Rather, the traditional methodologies like phonocardiogram, etc. are being utilized. The reason is the unavailability of an effective methodology for providing good quality fECG signal. The proposal of a hybrid tactic called Bayesian Deep Belief Network (BDBN) for fECG signal enhancement is presented in this article. The proposed BDBN technique involves Baye’s filtering methodology in amalgamation with Deep Belief Network. The Baye’s filtering was employed to eliminate undesired signal components. Deep learning (DL) technique was utilized with Deep belief network (DBN) to extract high quality fECG signal. The methodology resulted with good quality fECG signal which is indeed valuable for timely Physician analysis.     

Thoughts on representation in therapy of Holocaust survivors1

This paper presents the problems of representation and lack of representation in treating Holocaust survivors, through clinical vignettes and various theoreticians. The years of Nazi persecution and murder brought about a destruction of symbolization and turning inner and external reality into the Thing itself, the concrete, or, in Lacan’s words, ‘The Thing’. The paper presents two ideas related to praxis as well as theory in treating Holocaust survivors: the first is related to the therapist’s treatment of the Holocaust nightmare expressing the traumatic events just as they happened 63 years previously; the second deals with the attempt at subjectification, in contrast to the objectification forced by the Nazis on their victims.     

Préservation d’une image positive de soi dans un cadre de vie « à risque »

According to social identity theory (Tajfel, 1972), individual try to maintain a positive self-image. Places at which he/she is attached contribute to identity building (Proshansky et al., 1983; Twigger-Ross & Uzzell, 1996). Therefore, individual will try to maintain a positive a positive and stable image of such places. Our study is interested in what is the representation of their living area for individuals living in coastal area considered as exposed to erosion/flood risk, and attached to it. We are focused on how this representation influences their perception to coastal risk. We assume that individual attached to a place which he/she consider as “at risk” will put in place a strategy to preserve the positive image of this place and of himself/herself. Analyzing answers to a questionnaire, results demonstrate that individuals considering a place which they are attached as “at risk” mitigate the danger impact on that space. Thus, they conciliate probability of such negative event on place which they identifying themselves with positive image of this place and, by extension, of themselves.     

In which direction to move? Facilitative and interference effects of gestures on problem solver's thinking

Recent research shows that co-speech gestures can influence gesturers’ thought. This line of research suggests that the influence of gestures is so strong, that it can wash out and reverse an effect of learning. We argue that these findings need a more robust and ecologically valid test, which we provide in this article. Our results support the claim that gestures not only reflect information in our mental representations, but can also influence gesturer's thought by adding action information to one's mental representation during problem solving (Tower of Hanoi). We show, however, that the effect of gestures on subsequent performance is not as strong as previously suggested. As opposed to what previous research indicates, gestures' facilitative effect through learning was not nullified by the potentially interfering effect on subsequent problem-solving performance of incompatible gestures. To conclude, using gestures during problem solving seems to provide more benefits than costs for task performance.     

Excitatory and inhibitory priming by attended and ignored non-recycled words with monolinguals and bilinguals

Experiments examining identity priming from attended and ignored novel words (words that are used only once except when repetition is required due to experimental manipulation) in a lexical decision task are reported. Experiment 1 tested English monolinguals whereas Experiment 2 tested Twi (a native language of Ghana, Africa)-English bilinguals. Participants were presented with sequential pairs of stimuli composed of a prime followed by a probe, with each containing two items. The participants were required to name the target word in the prime display, and to make a lexical decision to the target item in the probe display. On attended repetition (AR) trials the probe target item was identical to the target word on the preceding attentional display. On ignored repetition (IR) trials the probe target item was the same as the distractor word in the preceding attentional display. The experiments produced facilitated (positive) priming in the AR trials and delayed (negative) priming in the IR trials. Significantly, the positive and negative priming effects also replicated across both monolingual and bilingual groups of participants, despite the fact that the bilinguals were responding to the task in their non-dominant language.     

Investigating Individual Differences in Toddler Search with Mixture Models

Children's performance on cognitive tasks is often described in categorical terms in that a child is described as either passing or failing a test, or knowing or not knowing some concept. We used binomial mixture models to determine whether individual children could be classified as passing or failing two search tasks, the DeLoache model room task and the Berthier et al. door task. The data support categorical classification of the children and suggest that the increase in average proportion correct with age is the result of an increasing proportion of children who can solve the tasks. Performance on the two tasks was concordant, and improving performance could be due to advances in a single psychological ability, such as cognitive control. Copyright © 2014 John Wiley & Sons, Ltd.     

A Single Counterexample Leads to Moral Belief Revision

What kind of evidence will lead people to revise their moral beliefs? Moral beliefs are often strongly held convictions, and existing research has shown that morality is rooted in emotion and socialization rather than deliberative reasoning. In addition, more general issues—such as confirmation bias—further impede coherent belief revision. Here, we explored a unique means for inducing belief revision. In two experiments, participants considered a moral dilemma in which an overwhelming majority of people judged that it was inappropriate to take action to maximize utility. Their judgments contradicted a utilitarian principle they otherwise strongly endorsed. Exposure to this scenario led participants to revise their belief in the utilitarian principle, and this revision persisted over several hours. This method provides a new avenue for inducing belief revision.     

Belief in a just world mediates the relationship between institutional trust and life satisfaction among the elderly in China

This study investigated the relationship between institutional trust and life satisfaction, and the mediating role of belief in a just world (BJW) among the elderly. The General Belief in a Just World Scale (GBJW) and Satisfaction with Life Scale (SWLS) were employed. A self-developed Institutional Trust questionnaire was used to measure participant’ levels of trust in eight institutions. The aggregate score for all eight items represented the level of institutional trust. The questionnaires were completed by 19,352 retirees ranging in age from 50 to 99 (M = 69.7, SD = 8.0). The results showed the following: (1) overall, the retirees tended to report high institutional trust and high life satisfaction; (2) institutional trust was positively associated with life satisfaction; and (3) more importantly, the relationships between institutional trust and life satisfaction were partially mediated by GBJW. This finding provides a new insight into the psychological mechanisms by which institutional trust relates to individual happiness. The theoretical and practical implications of these findings, as well as the study’s limitations, are discussed.