ILLC-LIRa June 2020

illc-lira@list.uva.nl

1 participants
8 discussions

LIRa session (online only): Thomas Bolander and Lasse Dissing
by LIRa Seminar 24 Jun '20

24 Jun '20

Dear all, We will have our next LIRa session tomorrow. We will use our recurring zoom link:: https://uva-live.zoom.us/j/92907704256?pwd=anY3WkFmQVhLZGhjT2JXMlhjQVl1dz09 (Meeting ID: 929 0770 4256, Password: 036024) You can find the details of the talk below and the guidelines for our online sessions here: https://projects.illc.uva.nl/lgc/seminar/guidelines-for-online-sessions/ Speaker: Thomas Bolander and Lasse Dissing Date and Time: Thursday, June 25th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Implementing Theory of Mind on a Robot Using Dynamic Epistemic Logic Abstract. Previous research has claimed dynamic epistemic logic (DEL) to be a suitable formalism for representing essential aspects of a Theory of Mind (ToM) for autonomous agents. This includes the ability of the formalism to represent the reasoning involved in false-belief tasks of arbitrary order, and hence for autonomous agents based on the formalism to become able to pass such tests. We provide evidence for the claims by presenting the implementation of a DEL-based reasoning system on a humanoid robot (the Pepper robot by Softbank Robotics). Our implementation allows the robot to perform cognitive perspective-taking, in particular to reason about the first- and higher-order beliefs of other agents. We demonstrate how this allows the robot to pass a quite general class of false-belief tasks involving human agents. Additionally, it allows the robot to proactively provide human agents with relevant information in situations where a system without ToM-abilities would fail. The symbolic grounding problem of turning robotic sensor input into logical action descriptions in DEL is achieved via a perception system based on deep neural networks. Hope to see you there! The LIRa team

1 0

LIRa session (online only): Thomas Bolander and Lasse Dissing
by LIRa Seminar 19 Jun '20

19 Jun '20

Dear all, We will have our next LIRa session on Thursday, June 25th. Our speakers are Thomas Bolander and Lasse Dissing. You can find the details of the talk below. As usual, the talk will be followed by (online) drinks. We will use our recurring zoom link:: https://uva-live.zoom.us/j/92907704256?pwd=anY3WkFmQVhLZGhjT2JXMlhjQVl1dz09 (Meeting ID: 929 0770 4256, Password: 036024) Speaker: Thomas Bolander and Lasse Dissing Date and Time: Thursday, June 25th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Implementing Theory of Mind on a Robot Using Dynamic Epistemic Logic Abstract. Previous research has claimed dynamic epistemic logic (DEL) to be a suitable formalism for representing essential aspects of a Theory of Mind (ToM) for autonomous agents. This includes the ability of the formalism to represent the reasoning involved in false-belief tasks of arbitrary order, and hence for autonomous agents based on the formalism to become able to pass such tests. We provide evidence for the claims by presenting the implementation of a DEL-based reasoning system on a humanoid robot (the Pepper robot by Softbank Robotics). Our implementation allows the robot to perform cognitive perspective-taking, in particular to reason about the first- and higher-order beliefs of other agents. We demonstrate how this allows the robot to pass a quite general class of false-belief tasks involving human agents. Additionally, it allows the robot to proactively provide human agents with relevant information in situations where a system without ToM-abilities would fail. The symbolic grounding problem of turning robotic sensor input into logical action descriptions in DEL is achieved via a perception system based on deep neural networks. Hope to see you there! The LIRa team

1 0

LIRa session (online only): Jörg Endrullis
by LIRa Seminar 17 Jun '20

17 Jun '20

1 0

LIRa session (online only): Jörg Endrullis
by LIRa Seminar 12 Jun '20

12 Jun '20

Dear all, We will have our next LIRa session on Thursday, June 18th. Our speaker is Jörg Endrullis. You can find the details of the talk below. As usual, the talk will be followed by (online) drinks. We will use our recurring zoom link:: https://uva-live.zoom.us/j/92907704256?pwd=anY3WkFmQVhLZGhjT2JXMlhjQVl1dz09 (Meeting ID: 929 0770 4256, Password: 036024) Speaker: Jörg Endrullis Date and Time: Thursday, June 18th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Syllogistic Logic with \"Most\" Abstract. We add Most X are Y to the syllogistic logic of All X are Y and Some X are Y. We prove soundness, completeness, and decidability in polynomial time. Our logic has infinitely many rules, and we prove that this is unavoidable. This is joint work with Larry Moss. Hope to see you there! The LIRa team

1 0

LIRa session (online only): Tobias Blanke
by LIRa Seminar 10 Jun '20

10 Jun '20

Dear all, Tomorrow we will have our next LIRa session. We will be using the following recurring zoom meeting: https://uva-live.zoom.us/j/92907704256?pwd=anY3WkFmQVhLZGhjT2JXMlhjQVl1dz09 (Meeting ID: 929 0770 4256, Password: 036024) You can find the details of the talk below and the guidelines for our online sessions here: https://projects.illc.uva.nl/lgc/seminar/guidelines-for-online-sessions/ Speaker: Tobias Blanke Date and Time: Thursday, June 11th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Where are the humanities in AI research? – a research agenda Abstract. The talk will try to ask the question where the humanities are in AI research. I will follow their historical position but mainly focus on their contemporary relevance with specific research questions and interests targeting developments in digital culture and society. The digital devices we develop need to focus on the critical study of society. This includes algorithmic criticism but also the study of platformisation through critical hacking, to name but a few. Hope to see you there! The LIRa team

1 0

LIRa session (online only): Tobias Blanke
by LIRa Seminar 05 Jun '20

05 Jun '20

Dear all, We will have our next LIRa session on Thursday, June 11th. Our speaker is Tobias Blanke. You can find the details of the talk below. As usual, the talk will be followed by (online) drinks. After a suggestion of the audience, we will use a recurring zoom meeting. This means that you can use the following link, not only for the session next week, but also for all remaining LIRa sessions: https://uva-live.zoom.us/j/92907704256?pwd=anY3WkFmQVhLZGhjT2JXMlhjQVl1dz09 (Meeting ID: 929 0770 4256, Password: 036024). In any case, we will send regular reminders as usual. Speaker: Tobias Blanke Date and Time: Thursday, June 11th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Where are the humanities in AI research? – a research agenda Abstract. The talk will try to ask the question where the humanities are in AI research. I will follow their historical position but mainly focus on their contemporary relevance with specific research questions and interests targeting developments in digital culture and society. The digital devices we develop need to focus on the critical study of society. This includes algorithmic criticism but also the study of platformisation through critical hacking, to name but a few. Hope to see you there! The LIRa team

1 0

LIRa session (online only): Adam Bjorndahl
by LIRa Seminar 03 Jun '20

03 Jun '20

Dear all, Tomorrow we will have our next LIRa session. Here are the zoom details: https://uva-live.zoom.us/j/99889363903?pwd=V1hMQ3pvaVdDMURUZ2Z1c085RGtadz09 Meeting ID: 998 8936 3903 Password: 741643 You can find the details of the talk below and the guidelines for our online sessions here: https://projects.illc.uva.nl/lgc/seminar/guidelines-for-online-sessions/ Speaker: Adam Bjorndahl Date and Time: Thursday, June 4th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Almost-logic Abstract. In standard possible worlds style semantics for modal logics, a model consists in a set of worlds W together with some additional structure (e.g., a relation, a topology, a (set of) function(s), etc.). And a formula is defined to be valid in such a model if it is true at each and every world in W. In this talk, we consider the idea of relaxing this definition of validity: instead of requiring truth at all worlds in W, what happens if we only ask for truth at “almost all” worlds? Of course, this depends on just what we mean by “almost all”. Natural closure conditions on the corresponding notion of “almost-validity” yield some constraints, but of course do not determine a unique definition of “almost all”. On the other hand, well-known topological and measure-theoretic notions of “large” sets (and, dually, “negligible” sets) provide appealing candidates for making this notion precise; each determines a corresponding class of “almost-valid” formulas, with some surprising and familiar axiomatizations to explore. Hope to see you there! The LIRa team

1 0

LIRa session (online only): Adam Bjorndahl
by LIRa Seminar 01 Jun '20

01 Jun '20

Dear all, We will have our next LIRa session on Thursday, June 4th. Our speaker is Adam Bjorndahl. You will receive the zoom link and a reminder of the guidelines for online sessions the day before the talk. After the talk, we will continue the discussion with (online) drinks, as usual. You can find the details below. Speaker: Adam Bjorndahl Date and Time: Thursday, June 4th 2020, 16:30-18:00, Amsterdam time. Venue: online. Title: Almost-logic Abstract. In standard possible worlds style semantics for modal logics, a model consists in a set of worlds W together with some additional structure (e.g., a relation, a topology, a (set of) function(s), etc.). And a formula is defined to be valid in such a model if it is true at each and every world in W. In this talk, we consider the idea of relaxing this definition of validity: instead of requiring truth at all worlds in W, what happens if we only ask for truth at “almost all” worlds? Of course, this depends on just what we mean by “almost all”. Natural closure conditions on the corresponding notion of “almost-validity” yield some constraints, but of course do not determine a unique definition of “almost all”. On the other hand, well-known topological and measure-theoretic notions of “large” sets (and, dually, “negligible” sets) provide appealing candidates for making this notion precise; each determines a corresponding class of “almost-valid” formulas, with some surprising and familiar axiomatizations to explore. Hope to see you there! The LIRa team

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

ILLC-LIRa June 2020