People
Who we are
Kangil Kim | Professor
Associate Professor,
School of Electrical Engineering & Computer Science,
Gwangju Institute of Science and Technology(GIST)
email: kangil.kim.01[at]gmail[dot]com / kikim01[at]gist[dot]ac[dot]kr
CV
Research Interest
generally applicable intelligence model representation, optimization and regularization, model complexity, probabilistic and representation analysis.
Research Area
artificial intelligence, machine learning, evolutionary computation, neural network, probabilistic graphical model, statistical relation learning, logic and grammar model, genetic programming, natural language processing
Education
2012 Ph.D in Computer Science and Engineering, SNU
2006 B.S. in Computer Science, KAIST
Experience
2022~present Associate Professor, GIST, AI Graduate School
2019~2022 Assistant Professor, GIST, EECS
2016~2019 Assistant Professor, KU, CSE
2013~2016 Senior Researcher, ETRI, AI and NLP
2012~2013 Postdoc Researcher, SNU, Structural Complexity Lab
2011 Visiting Researcher, UPM, Artificial Intelligence Group
2008 Research Intern, NII
Publications - Selected
[TMLR] H.-J. Jung, J. Jeong, and K. Kim, “CFASL: Composite Factor-Aligned Symmetry Learning for Disentanglement in Variational AutoEncoder”, 2024
[ACL Findings] J. Park, and K. Kim., “Structural Optimization Ambiguity and Simplicity Bias in Unsupervised Neural Grammar Induction”, 2024
[ICLR] I. Kang, H. Bae, and K. Kim. “Label-Focused Inductive Bias over Latent Object Features in Visual Classification”, 2024
[ICLR] H. Kim, and K. Kim. “Fixed Non-negative Orthogonal Classifier: Inducing Zero-mean Neural Collapse with Feature Dimension Separation”, 2024
[CoLLAS (workshop)] Asymptotic Midpoint Mixup for Margin Balancing and Moderate Broadening, H Kim, S Lee, K Kim, 2024
[CoLLAS (workshop)] Revisiting Softmax Masking: Stop Gradient for Enhancing Stability in Replay-based Continual Learning, H Kim, M Kwon, K Kim, 2024
[arxiv] Multiple Invertible and Equivariant Transformation for Disentanglement in VAEs, HJ Jung, K Kim, 2024
[arxiv] Enhancing Accuracy and Robustness through Adversarial Training in Class Incremental Continual Learning, M Kwon, K Kim, 2023
[ESWA] H.-J. Jung, D. Kim, S.-H. Na, and K. Kim. Feature structure distillation for bert transferring. 2023
[NeuRIPS] H. Kim, and K. Kim. “Spherization Layer: Representation Using Only Angles”, NeurIPS 2022.
[IEEE TNNLS] S. Woo, J. Noh, and K. Kim. Tackling the challenges in scene graph generation with local-to-global interactions. 2022.
[arxiv] S. Woo, J. Noh, and K. Kim. What and when to look?: Temporal span proposal network for video visual relation detection. arXiv preprint arXiv:2107.07154, 2021.
[ESWA] K. Kim, Y. Jin, S. Nah, and Y.-K. Kim. Center-shared sliding ensemble of neural networks for syntax analysis of natural language. Expert Systems with Applications, 83:215–225, 2017.
[IEEE TEVC] K. Kim, R. B. McKay, and N. X. Hoai. Recursion-based biases in stochastic grammar model genetic programming. 2016.
[GPEM] K. Kim, Y. Shan, X. H. Nguyen, and R. I. McKay. Probabilistic model building in genetic programming: a critical review. Genetic Programming and Evolvable Machines, 15(2):115–167, 2014.
[IEEE TEVC] K. Kim and R. I. McKay. Stochastic diversity loss and scalability in estimation of distribution genetic programming. 2013.
- (all) links: google scholar, research gate
Main Projects
- Model Analysis
2024~2025 Theoretical Stagnation of Foundation Models
2022~2027 Representation Analysis for Learning Canonicalized Representations (NRF)
2019~2021 Neural Networks for General Knowledge Accummulation (GIST, NRF)
2016~2016 Improvement of Expression Power of Infinite Probabilistic Context Free Grammar (NRF)
before 2013 Bias analysis of probabilistic context free grammar models in estimation of distribution algorithm in genetic programming (in PhD program, NRF)
2019~2021 Neural Networks for General Knowledge Accummulation (GIST, NRF) - Natural Language Processing
2016~2020 Real-time multilingual machine translation (ETRI)
2013~2016 Multilingual machine translation (ETRI)
Review Service
IEEE Transactions on Neural Network and Learning Systems
IEEE Transactions on Evolutionary Computation
IEEE Transactions on Audio, Speech and Language Processing
Expert Systems With Applications
NeuRIPS, ICLR, ICML, ACL, EMNLP
Acronyms
GIST (Gwangju Institute of Science and Technology, Korea)
ETRI (Electronics and Telecommunications Research Institute, Korea)
UPM (Polytechnic University of Madrid, Spain)
SNU (Seoul National University, Korea)
NII (National Institute of Informatics, Japan)