KAKENHI meeting FY2023

Program

# Thursday (September 14th) Friday (September 15th)
10:00

Opening & Announcements

:10
:20

On Causal Prevention

Katsumi Inoue

:30

Counterfactual explanation from (neuro-)symbolic view

Tony Ribeiro

:40

A formal account of Hollis's paradox

Chiaki Sakama

:50
11:00

Progress report of T-PRISM

Ryosuke Kojima

:10
:20
:30

Parsimonious Equation Learning with Causality

Mitsuhiro Odaka

Class-Incremental Learning using Diffusion Model for Distillation and Replay

Phua Yin Jun

:40
:50
12:00

Lunch

:10

Lunch

:20
:30
:40
:50
13:00
:10
:20
:30

Preference-based Repairs of Databases and Knowledge Bases

Meghyn Bienvenu

:40

Discovering Relations between Neural State Variables for Symbolic Regression

Koji Watanabe

:50
14:00
:10
:20
:30

Understanding Symbol Binding in Multimodal Transformers

Alex Spies

Network construction via graph neural cellular automata

Xu Fanqing

:40
:50
15:00
:10

BeliefFlow: a Framework for Logic-Based Belief Diffusion via Iterated Belief Change

Nicolas Schwind

Break

:20
:30

When ML gets wrong... KR to the rescue!

Pierre Marquis

:40

Break

:50
16:00

DNF as a neurral network and interpolant learning

Taisuke Sato

:10
:20
:30

Discussion & Closing

:40

Differentiable solver for Sudoku and Logic Programs under Stable Model Semantics

Akihiro Takemura

:50
17:00
:10

From stable model condition to learning normal rules in the form of program matrices

Tuan Nguyen

:20
:30
:40
:50
Caption
Morning Session (Day1)
Afternoon Session 1 (Day1)
Afternoon Session 2 (Day1)
Morning Session (Day2)
Afternoon Session 1 (Day2)
Afternoon Session 2 (Day2)
Discussion & Closing
Break