536 research outputs found
How Can We Make Trustworthy AI? (Invited Talk)
No full textNot too long ago most headlines talked about our fear of AI. Today, AI is ubiquitous, and the conversation has moved on from whether we should use AI to how we can trust the AI systems that we use in our daily lives. In this talk I look at some key technical ingredients that help us build confidence and trust in using intelligent technology. I argue that intuitiveness, interaction, explainability and inclusion of human domain knowledge are essential in building this trust. I present some of the techniques and methods we are building for making AI systems that think and interact with humans in more intuitive and personalised ways, enabling humans to better understand the solutions produced by machines, and enabling machines to incorporate human domain knowledge in their reasoning and learning processes
. . . . . .
No full textions in Diagrams Mateja Jamnik Course Suitability: Any Languages and/or other Software: Any (an AI language such as Prolog, Lisp or ML might be well suited). Prerequisite: Good knowledge of mathematics and logic is required. 1 Abstractions in Diagrams Abstraction devices, such as ellipsis, are conventions and notations which are used to represent generality or abstraction of a structure. They can be used in sentential (symbolic) reasoning (e.g. n 2 = 1 + 3 + \Delta \Delta \Delta + (2n \Gamma 1)) or in diagrammatic reasoning. For example, representing the most general representation of a square of a natural size n requires the use of ellipsis in the diagram representation. Figure 1 shows the representation of an abstract square. . . . . . . . . . . . . . . . . . . Figure 1: Abstract representations of a square. The fact that diagrams are concrete in nature is an inherent problem in drawing diagrams. Using ellipsis is one of the conventions to represent generality. However, t..
A Graphical User Interface Framework for Formal Verification
No full textWe present the "ProofWidgets" framework for implementing general user interfaces (UIs) within an interactive theorem prover. The framework uses web technology and functional reactive programming, as well as metaprogramming features of advanced interactive theorem proving (ITP) systems to allow users to create arbitrary interactive UIs for representing the goal state. Users of the framework can create GUIs declaratively within the ITP’s metaprogramming language, without having to develop in multiple languages and without coordinated changes across multiple projects, which improves development time for new designs of UI. The ProofWidgets framework also allows UIs to make use of the full context of the theorem prover and the specialised libraries that ITPs offer, such as methods for dealing with expressions and tactics. The framework includes an extensible structured pretty-printing engine that enables advanced interaction with expressions such as interactive term rewriting. We exemplify the framework with an implementation for the https://leanprover-community.github.io. The framework is already in use by hundreds of contributors to the Lean mathematical library
Learning Strategies for Mechanised Building of Decision Procedures
No full textAbstractIn this paper we present an investigation into whether and how decision procedures can be learnt and built automatically. Our approach consists of two stages. First, a refined brute-force search procedure applies exhaustively a set of given elementary methods to try to solve a corpus of conjectures generated by a stochastic context-free grammar. The successful proof traces are saved. In the second stage, a learning algorithm (by Jamnik et al.) tries to extract a required supermethod (i.e., decision procedure) from the given traces. In the paper, this technique is applied to elementary methods that encode the operations of the Fourier-Motzkin's decision procedure for Presburger arithmetic on rational numbers. The results of our experiment are encouraging
Work on the role of Adder in the play "Striborov's Forest"
No full textŠuma Striborova je diplomska predstava četvero studenata: Laure Kolesarić, Luke Selmana, Matka Trnačića i Mateje Tustanovski pod mentorstvom Maje Lučić, izv. prof. dr. art., i sumentorstvom umjetničkog suradnika Gorana Guksića. Mateja Tustanovski ovim radom dramaturški analizira komad, obrazlaže i argumentira izbor tehnike te opisuje svoj proces rada na ulozi Guje kao i cjelokupnoj predstavi. Na kraju autorica donosi zaključak o stečenom iskustvu tijekom procesa.„Stribor's forrest“ is a graduation thesis of four students, including Laura Kolesarić, Luka Selman, Matko Trnačić and Mateja Tustanovski, under the mentorship of izv. prof. ArtD. Maja 24 Lučić and co-mentorship of ass. Goran Guksić umj. sur. With this thesis, Mateja Tustanovski explains and argues the choice of technique, dramaturgically analyzes the play, and describes her process of the work on the role and the play. At the end, the author draws the conclusion about the experience gained in the process
Work on the role of Adder in the play "Striborov's Forest"
No full textŠuma Striborova je diplomska predstava četvero studenata: Laure Kolesarić, Luke Selmana, Matka Trnačića i Mateje Tustanovski pod mentorstvom Maje Lučić, izv. prof. dr. art., i sumentorstvom umjetničkog suradnika Gorana Guksića. Mateja Tustanovski ovim radom dramaturški analizira komad, obrazlaže i argumentira izbor tehnike te opisuje svoj proces rada na ulozi Guje kao i cjelokupnoj predstavi. Na kraju autorica donosi zaključak o stečenom iskustvu tijekom procesa.„Stribor's forrest“ is a graduation thesis of four students, including Laura Kolesarić, Luka Selman, Matko Trnačić and Mateja Tustanovski, under the mentorship of izv. prof. ArtD. Maja 24 Lučić and co-mentorship of ass. Goran Guksić umj. sur. With this thesis, Mateja Tustanovski explains and argues the choice of technique, dramaturgically analyzes the play, and describes her process of the work on the role and the play. At the end, the author draws the conclusion about the experience gained in the process
Workshop on Human-Interpretable AI
No full textThis workshop aims to spearhead research on Human-Interpretable Artificial Intelligence (HI-AI) by providing: (i) a general overview of the key aspects of HI-AI, in order to equip all researchers with the necessary background and set of definitions; (ii) novel and interesting ideas coming from both invited talks and top paper contributions; (iii) the chance to engage in dialogue with prominent scientists during poster presentations and coffee breaks. The workshop welcomes contributions covering novel interpretable-by-design or post-hoc approaches, as well as theoretical analysis of existing works. Additionally, we accept visionary contributions speculating on the future potential of this field. Finally, we welcome contributions from related fields such as Ethical AI, Knowledge-driven Machine learning, Human-machine Interaction, but also applications in Medicine and Industry, and analyses from Regulatory experts
Concept Distillation in Graph Neural Networks
No full textThe opaque reasoning of Graph Neural Networks induces a lack of human trust. Existing graph network explainers attempt to address this issue by providing post-hoc explanations, however, they fail to make the model itself more interpretable. To fill this gap, we intro- duce the Concept Distillation Module, the first differentiable concept- distillation approach for graph networks. The proposed approach is a layer that can be plugged into any graph network to make it explainable by design, by first distilling graph concepts from the latent space and then using these to solve the task. Our results demonstrate that this approach allows graph networks to: (i) attain model accuracy comparable with their equivalent vanilla versions, (ii) distill meaningful concepts achiev- ing 4.8% higher concept completeness and 36.5% lower purity scores on average, (iii) provide high-quality concept-based logic explanations for their prediction, and (iv) support effective interventions at test time: these can increase human trust as well as improve model performance
Carniolan and Styrian identity in Posavje
No full textV magistrskem delu se avtorica ukvarja z raziskovanjem lokalnega pojava v Posavju, natančneje z aktualizacijo uporabe deželne identitete. V teoretičnem delu je pozornost sprva namenjena osnovnim pojmom, ki so povezani s tematiko. V nadaljevanju sta izpostavljeni dve zgodovinski dejstvi za lažje razumevanje raziskovalnega vprašanja, in sicer do kdaj so upravno obstajale dežele in kakšno je stanje na področju regionalizacije po osamosvojitvi Slovenije. Nato se je avtorica osredotočila na različne (neupravne) regionalizacije, znotraj katerih je umestila tudi raziskovalni prostor – Posavje. V okviru teorije je govora prav tako še o meji ter identiteti. Na koncu teoretičnega dela se je avtorica osredotočila konkretno še na deželno identiteto v Posavju. Medtem ko glavni del magistrskega dela predstavlja analiza aktualizacije uporabe deželne identitete v Posavju. V analizi se je avtorica osredotočila na identificiranje ljudi z deželno identiteto nasplohkdo in kakšna je oseba z deželno identiteto oz. kaj je zanjo značilnokakšni so razlogi za uporabo deželne identitete pri ljudehkako se deželna identiteta uporablja in da se uporablja predvsem na način šaljenjakakšna čustva se pojavijo pri uporabi deželne identitetekje se pojavi deželna identiteta v družabnem življenju in kaj je lahko v pomoč pri izražanju deželne identitete. V magistrskem delu se je za posebno pomembno pokazala ena ugotovitev, za katero avtorica lahko reče, da je kar glavna ugotovitev raziskovanja, in sicer da je za uporabo deželne identitete značilno šaljenje oz. da se uporablja na šaljiv način. Drugače pa je avtorica z analiziranim gradivom in smernimi vprašalnicami pokazala oz. dokazala, da se v Posavju aktualizira uporaba deželne identitete.In the master\u27s thesis the author deals with the research of a local phenomenon in Posavje, the so-called modernisation of state identity use. The theoretical part deals with basic terms connected with the main topic. In the following chapters stress two historical facts to understand the research question easier – how long state administration has been present and what is the situation in the field of regionalisation after Slovenian\u27s independence. Furthermore, the author focuses on different (non-administrative) regionalisations within which the research area of Posavje is placed. The theory explains also border and identity. The theoretical part ends with focus on state identity in Posavje. The main part of the thesis, on the other hand, presents the analysis of state identity use modernisation in Posavje. The author deals with identification of people with state identity in general, who is or what characteristics such a person has, what is typical of this person, what are the reasons to use state identity among people, how state identity is used and when teasing is used, what emotions are present in state identity use, where state identity is used in social life and what kind of help is used when expressing state identity. The master\u27s thesis has pointed out one vital finding which is treated by the author as the most important research finding – teasing or joking is used when state identity takes place. The author has concluded that state identity use is modernized in Posavje which has been supported by material analyzed and guided questions
- …
