730 research outputs found
Stepwise Verification and Remediation of Student Reasoning Errors with Large Language Model Tutors
No full textLarge language models (LLMs) present an opportunity to scale high-quality personalized education to all. A promising approach towards this means is to build dialog tutoring models that scaffold students\u27 problem-solving. However, even though existing LLMs perform well in solving reasoning questions, they struggle to precisely detect student\u27s errors and tailor their feedback to these errors. Inspired by real-world teaching practice where teachers identify student errors and customize their response based on them, we focus on verifying student solutions and show how grounding to such verification improves the overall quality of tutor response generation. We collect a dataset of 1K stepwise math reasoning chains with the first error step annotated by teachers. We show empirically that finding the mistake in a student solution is challenging for current models. We propose and evaluate several verifiers for detecting these errors. Using both automatic and human evaluation we show that the student solution verifiers steer the generation model towards highly targeted responses to student errors which are more often correct with less hallucinations compared to existing baselines.Preprint. Nico Daheim and Jakub Macina contributed equally. Code and dataset can be found under: https://github.com/eth-lre/verify-then-generat
MathDial: A Dialogue Tutoring Dataset with Rich Pedagogical Properties Grounded in Math Reasoning Problems
No full textWhile automatic dialogue tutors hold great potential in making education
personalized and more accessible, research on such systems has been hampered by
a lack of sufficiently large and high-quality datasets. Collecting such
datasets remains challenging, as recording tutoring sessions raises privacy
concerns and crowdsourcing leads to insufficient data quality. To address this,
we propose a framework to generate such dialogues by pairing human teachers
with a Large Language Model (LLM) prompted to represent common student errors.
We describe how we use this framework to collect MathDial, a dataset of 3k
one-to-one teacher-student tutoring dialogues grounded in multi-step math
reasoning problems. While models like GPT-3 are good problem solvers, they fail
at tutoring because they generate factually incorrect feedback or are prone to
revealing solutions to students too early. To overcome this, we let teachers
provide learning opportunities to students by guiding them using various
scaffolding questions according to a taxonomy of teacher moves. We demonstrate
MathDial and its extensive annotations can be used to finetune models to be
more effective tutors (and not just solvers). We confirm this by automatic and
human evaluation, notably in an interactive setting that measures the trade-off
between student solving success and telling solutions. The dataset is released
publicly.Comment: Jakub Macina, Nico Daheim, and Sankalan Pal Chowdhury contributed
equally to this work. Accepted at EMNLP2023 Findings. Code and dataset
available: https://github.com/eth-nlped/mathdia
Treatment of unruptured intracranial aneurysms : a surgical long-term evaluation for preoperative predictive analytics
No full textAuthor Dr.med.univ. Nico Henrique Stroh-HollyDissertation Johannes Kepler Universität Linz 202
Enough of 'tough': Youth Justice in Scotland
No full textHow can we build an effective youth justice system that enjoys public confidence while recognising that children must be redeemable?Nico Juetten presents a view from Scotland, considering the challenges to the children's hearings system Copyright (c) 2009 The Author. Journal compilation (c) 2009 ippr.
Ciò che resta della terza dimensione
No full textThe s.c. «third dimension» is one of the first definitions given in the sociological arena to define the development of a ate element that was not identifiable with generalist instances of the state or with those particularistic of the market. Deriving its conceptualization from the Husserlian theory, Ardigò through the use this approach emphasizing the need to reconstruct the crisis through new forms of transaction between the social system and life-worlds. The use of this conceptualization led to the identification of these intermediary organizations between different systems. The text takes a brief overview of the third dimension organizations and the way how they operate the reconciliation between inter-subjectivity and super-ordination present in the systemic sense. It is proposed a scheme of interpretation, drawn from the author in question, through which classifying not merely the organization but the possible interconnections between micro and macro and their consequences. It is concluded that the abandon of the analytical perspective proposed by Ardigò, primarily for the affirmation of a vision of an economic nature, was determined both by the theoretical vulgate established around of the 90, and the difficulty in operationalizing a definition that owed much to the knowledge from the phenomenology of the Bolognese scientist
Alla ricerca della sovranità nello studio di Carlo Lavagna sulle figure giuridiche soggettive
No full textIl presente contributo ripercorre il saggio di Carlo Lavagna, Basi per uno studio delle figure giuridiche soggettive contenute nella Costituzione italiana, allo scopo di ricercare il filo conduttore della sovranità in tutte le figure giuridiche soggettive sia che esse siano solo raccomandate, sia che siano garantite o anche tutelate dalla Costituzione, secondo la classificazione dell’Autore. Tale analisi è stata preceduta da un breve richiamo al metodo di indagine dell’Autore e alla sua concezione quadrangolare della democrazia.This contribution reinterprets the essay by Carlo Lavagna, Basis for a study of the subjective juridical figures contained in the Italian Constitution, in order to search for the sphere of sovereignty in all subjective legal figures whether they are only recommended, whether they are guaranteed or also protected by the Constitution, according to the classification of the Author. This analysis was preceded by a brief recall of the Author's method of investigation and its quadrangular concept of democracy
The Understood Author: A hermeneutical exploration of audiences interpretation of the author as productive practices behind a text
No full textin English This project builds on the tradition of audience research that theoretically and empirically established that meaning is formed in the negotiation between the text and the reader, that the audiences are agentic, their interpretation of media is contextual and situated, and that they have the potential to resist the media. It argues that it is important to explore people's attention to different sources: who is listening, to whom, and why - and asks how people understand the notion of an author that they encounter in the interpretation of factual media texts, and how this interpretation is brought back into the interpretative act. The thesis subscribes to philosophical hermeneutics as an overarching theoretical as well as methodological framework to devise a concept of an understood author that is a result of the interpretative act, where the author imagined and anticipated by the reader (in the form of prior knowledge and prejudices) is encountered and actualised by the author inscribed into the text. I carried out 28 in-depth face-to-face interviews with Czechs born between 1948-1963 to explore participants' experiences and relationship with media and media authors. Employing hermeneutical analysis, the focus is not to capture what is an author, but how it happened to be so. The..
NILAI NILAI MORAL DALAM FILM NOBODY KNOWS KARYA SUTRADARA HIROKAZU KOREEDA KAJIAN SOSIOLOGI SASTRA 『是枝 裕和 が監督した『NOBODY KNOWS 』という映画における 道徳的価値『文芸社会学の研究』
No full textABSTRACT
Saragih, Nico Rinaldi. 2018. "Moral Values in the Nobody Knows Movie by Director of Hirokazu Koreeda in the Study of Sociology of Literature". A Thesis of Japanese Language and Culture, Faculty of Humanities Diponegoro University, Semarang. Consultant: Budi Mulyadi, S.Pd.,M.Hum.
This study aims to describe the narrative element of the film Nobody Knows and describe the moral values contained in the film Nobody Knows. The author uses the literature study method to obtain data and also data observations by viewing the Nobody Knows movie over and over. The theory used in this study is moral value by Andri Wicaksono.
Based on data analysis, it can be concluded that in the Nobody Knows film there are 8 moral values contained in human and self relations, and 5 moral values contained in human relations with other human beings in the social sphere
Keywords: Hirokazu Koreeda, Moral Value, Nobody Know
What's on a machine's mind? Models for reasoning with incomplete and uncertain knowledge
No full textElectrical Engineering, Mathematics and Computer Scienc
Vocabulary games: More than just wordplay
No full textGames offer more than just fun and play, and the analyses suggest that varying vocabulary activities is important if learners are to practice all the aspects of word knowledge. However, despite all the evidence, just claiming that playing games is a good way to practice the language is often not enough to win the argument. Because author Nico Lorenzutti's in-service teachers needed a specific rationale to explain what their learners achieve when they play games, he created a simple tool to explain what learners achieve when they play games. This article describes the procedures for several games and illustrates how to use this tool to assess their overall effectiveness for vocabulary development
- …
