Masahiro Negishi 根岸 優大

Research output

Publications

2026

IOP-MLST journal

Continuous SUN (Stable, Unique, and Novel) Metric for Generative Modeling of Inorganic Crystals

Masahiro Negishi, Hyunsoo Park, Kinga O. Mastej, Aron Walsh

IOPscience Machine Learning: Science and Technology

Paper

2025

NeurIPS conference

Continuous Uniqueness and Novelty Metrics for Generative Modeling of Inorganic Crystals

Masahiro Negishi, Hyunsoo Park, Kinga O. Mastej, Aron Walsh

AI for Accelerated Materials Design - NeurIPS 2025

Paper
ICML conference

WILTing Trees: Interpreting the Distance Between MPNN Embeddings

Masahiro Negishi, Thomas Gartner, Pascal Welke

ICML 2025

Paper
thesis

Weakly Supervised Disentanglement from Distance-based Supervision

Masahiro Negishi, Masashi Sugiyama

Master Thesis, Department of Information Science, School of Science, The University of Tokyo

2024

NeurIPS conference

Two-Stage Coefficient Estimation in Symbolic Regression for Scientific Discovery

Masahiro Negishi, Ryo Igarashi, Yoshitaka Ushiku, Yoshitomo Matsubara, Naoya Chiba

Machine Learning and the Physical Sciences Workshop - NeurIPS 2024

Paper

2023

JSAI journal

Scaling Laws of Dataset Size for VideoGPT

Masahiro Negishi, Makoto Sato, Ryosuke Unno, Koudai Tabata, Taiju Watanabe, Junnosuke Kamohara, Taiga Kume, Ryo Okada, Yusuke Iwasawa, Yutaka Matsuo

Proceedings of the Annual Conference of JSAI

Paper
JSAI journal

Construction and Validation of Action-Conditioned VideoGPT

Koudai Tabata, Junnosuke Kamohara, Ryosuke Unno, Makoto Sato, Taiju Watanabe, Taiga Kume, Masahiro Negishi, Ryo Okada, Yusuke Iwasawa, Yutaka Matsuo

Proceedings of the Annual Conference of JSAI

Paper
JSAI journal

Scaling Laws of Model Size for World Models

Makoto Sato, Ryosuke Unno, Masahiro Negishi, Koudai Tabata, Taiju Watanabe, Junnosuke Kamohara, Taiga Kume, Ryo Okada, Yusuke Iwasawa, Yutaka Matsuo

Proceedings of the Annual Conference of JSAI

Paper
thesis

Pairwise-constraint Classification in Weakly Supervised Machine Learning: Risk-consistent Approach and Classifier-consistent Approach

Masahiro Negishi, Masashi Sugiyama

Bachelor Thesis, Department of Information Science, School of Science, The University of Tokyo