Updated on 2026/07/28

写真a

 
phua yin jun
 
Organization
School of Computing Assistant Professor
Title
Assistant Professor
Homepage
https://yinjunphua.com/
External link

Degree

  • PhD (Informatics) ( 2022.3   The Graduate University for Advanced Studies )

Research Interests

  • Neuro-Symbolic

  • Logical Reasoning

  • Deep Learning

Research Areas

  • Informatics / Computational science

  • Informatics / Intelligent informatics

Education

  • The Graduate University for Advanced Studies   School of Multidisciplinary Sciences   Department of Informatics

    2019.4 - 2022.3

    　 More details

    Notes： PhD (Informatics)

    researchmap

  • Tokyo Institute of Technology   School of Computing   Department of Informatics

    2017.4 - 2019.3

    　 More details

  • Tokyo Institute of Technology   School of Engineering   Department of Computer Science

    2013.4 - 2017.3

    　 More details

Research History

  • Institute of Science Tokyo   School of Computing   Assistant Professor

    2024.10

    　 More details

  • Tokyo Institute of Technology   School of Computing   Assistant Professor

    2022.4 - 2024.9

    　 More details

  • Japan Society for the Promotion of Science   Research fellowship (DC2)

    2021.4 - 2022.3

    　 More details

  • National Institute of Informatics   Special RA

    2019.4 - 2021.3

    　 More details

Committee Memberships

  • International Conference on Neurosymbolic Learning and Reasoning   Program Committee  

    2024   

    　 More details

    Committee type：Academic society

    researchmap

Papers

▼display all

MISC

  • Can Transformers Learn to Verify During Backtracking Search? International coauthorship International journal

    Yin Jun Phua, Tony Ribeiro, Tuan Nguyen, Katsumi Inoue

    CoRR   abs/2605.22221   2026.5

    　More details

    Language：English   Publishing type：Rapid communication, short report, research note, etc. (scientific journal)  

    DOI： 10.48550/arXiv.2605.22221

    researchmap

  • Learning dynamics from limited training data with multilayer perceptron initialized by weight prediction

    PHUA Yin Jun, INOUE Katsumi

    Proceedings of the Annual Conference of JSAI   JSAI2018   4A101 - 4A101   2018

    　More details

    Language：Japanese   Publisher：The Japanese Society for Artificial Intelligence  

    Real world data are often difficult to obtain. Logical machine learning methods can produce perfect explanations for dynamics of systems when the full state transitions can be observed, but such scenario is often impossible. Statistical machine learning methods also usually require a huge amount of data. In this work, we propose a method that predicts the initial weight of an MLP to learn a model that can predict future state of a delayed system even when only a limited amount of observation is provided. We also show the effectiveness of the method applied to systems with particularly a large number of variables.

    DOI： 10.11517/pjsai.jsai2018.0_4a101

    CiNii Research

    researchmap

  • リカレントニューラルネットワークによる遅延を伴う解釈遷移から の論理プログラム表現学習

    ポア インジュン, 井上 克巳

    人工知能学会全国大会論文集   JSAI2017   3O11 - 3O11   2017

    　More details

    Language：Japanese   Publisher：一般社団法人 人工知能学会  

    Having a method to understand the interactions and delayed influences between components of dynamical systems can provide useful applications to biological and other dynamical systems. In this paper, we present a method relying on Recurrent Neural Networks (RNN) that can learn to distinguish the nature of different systems. This method utilizes Long Short-Term Memory (LSTM) to extract and encode certain features from the input sequence of time-series data. We also show that the produced high dimensional encoding can be used to represent different time series that are resulted from the same dynamical system.

    DOI： 10.11517/pjsai.jsai2017.0_3o11

    CiNii Research

    researchmap

Presentations

  • A Foundation Model for Learning Propositional Logic Program International conference

    Yin Jun Phua

    Ninth International Workshop on Symbolic-Neural Learning  2025.10 

    　More details

    Presentation type：Poster presentation  

    researchmap

  • Transformers Can Admit Mistakes and Backtrack International conference

    Tony Ribeiro, Yin Jun Phua, Tuan Nguyen, Katsumi Inoue

    2025.9 

    　More details

  • Leveraging Symbolic Invariance in Neuro-Symbolic AI Invited

    Yin Jun Phua

    2025.8 

    　More details

    Presentation type：Oral presentation (invited, special)  

    researchmap

  • Variable Assignment Invariant Neural Network for Learning Logic Programs International conference

    Yin Jun Phua, Katsumi Inoue

    2024.6 

    　More details

  • Learning from Noisy Transition Data International conference

    Yin Jun Phua, Katsumi Inoue

    2019.7 

    　More details

Works

Awards

  • Tokyo Institute of Technology Best Teacher Award

    2023.11   Tokyo Institute of Technology   Procedural Programming

    Masahito Ohue, Shio Miyafuji, Yin Jun Phua

    　More details

  • Best Paper Award

    2023.10   1st Workshop on Visual Continual Learning, ICCV2023   Class-Incremental Learning using Diffusion Model for Distillation and Replay

    Quentin Jodele, Xin Liu, Yin Jun Phua, Tsuyoshi Murata

    　More details

  • Best Student Paper Award

    2019.11   ILP 2019: 29th International Conference on Inductive Logic Programming   Learning Logic Programs from Noisy State Transition Data

    Yin Jun Phua, Katsumi Inoue

    　More details

Research Projects

  • Towards Reliable Generative AI Reasoning by Separated Search State Control Mechanism

    Grant number：262S08-24672  2026.7 - 2027.3

    National Institute of Informatics  National Institute of Informatics Open Collaborative Research 

    　 More details

    Authorship：Principal investigator 

    researchmap

  • Constraint Reasoning for Trustworthy AI

    Grant number：25K03190  2025.4 - 2029.3

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research  Grant-in-Aid for Scientific Research (B)

    Katsumi Inoue, Yuichi Sei, Hidetomo Nabeshima, Phua Yin Jun

    　 More details

    Grant type：Competitive

    researchmap

  • 背景知識と少量のデータから汎用的で解釈可能な知識をデータ駆動で学習するAIの開発

    Grant number：25K21269  2025.4 - 2028.3

    日本学術振興会  科学研究費助成事業  若手研究

    Phua Yin・Jun

    　 More details

    Grant amount：\4680000 （ Direct Cost: \3600000 、 Indirect Cost：\1080000 ）

    researchmap

  • Towards Reliable Generative AI by Logical Reasoning

    Grant number：24S1203  2024.7 - 2025.3

    National Institute of Informatics  National Institute of Informatics Open Collaborative Research 

    　 More details

    Authorship：Principal investigator 

    researchmap

  • Discovering New Knowledge by Combining Symbolic Logic and Deep Learning

    Grant number：22K21302  2022.8 - 2024.3

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research Grant-in-Aid for Research Activity Start-up  Grant-in-Aid for Research Activity Start-up

    　 More details

    Grant amount：\2860000 （ Direct Cost: \2200000 、 Indirect Cost：\660000 ）

    researchmap

  • Robust AI by Integration of Knowledge Representation and Machine Learning

    Grant number：21H04905  2021.4 - 2025.3

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research  Grant-in-Aid for Scientific Research (A)

    　 More details

    Grant amount：\41470000 （ Direct Cost: \31900000 、 Indirect Cost：\9570000 ）

    researchmap

  • Understandable Model by Combination of Symbolic Machine Learning and Deep Statistical Learning

    Grant number：21J14367  2021.4 - 2023.3

    Japan Society for the Promotion of Science  Grants-in-Aid for Scientific Research Grant-in-Aid for JSPS Fellows  Grant-in-Aid for JSPS Fellows

    Yin Jun Phua

    　 More details

    Grant amount：\1500000 （ Direct Cost: \1500000 ）

    researchmap

▼display all

Teaching Experience

  • Advanced Procedural Programming

    2023.6

    　More details