Tatsuro Inaba (稲葉達郎)
Tatsuro Inaba is a first-year Ph.D. student in Tohoku NLP Group at Tohoku University. He also serves as a research assistant at Research and Development Center for Large Language Models, National Institute of Informatics. His research interests lie at the intersection of language and music, with a particular focus on the internal mechanisms of Neural Networks.
- Email: tatsuro.pianooo[at]gmail.com
- Twitter(X)
- Google Scholar
- GitHub
News
- 06/2025: Gave a talk and tutorials (Pytorch/Transformer/Music Generation) at Kyushu University. [talk slides]
- 04/2025: Joined Tohoku NLP Group as a Ph.D. student.
- 03/2025: First-author preprint "How LLMs Learn: Tracing Internal Representations with Sparse Autoencoders." posted on arXiv. [arXiv].
Career
Education
- 04/2025-present: Ph.D. student, Graduate School of Information Sciences, Tohoku University.
- 04/2023-03/2025: Master student, Graduate School of Informatics, Kyoto University.
- 04/2019-03/2023: Bachelor student, Undergraduate School of Electrical and Electronic Engineering, Kyoto University.
Experience
- 12/2024-present: Research Assistant, Research and Development Center for Large Language Models, National Institute of Informatics.
- 01/2024-02/2024: Visiting Student, Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI).
- 11/2023-11/2024: Research Assistant, Kyoto University.
- 09/2023-10/2023: Machine Learning Engineer, Internship, Recruit Co., Ltd.
- 08/2023-09/2023: Researcher, Internship, Preferred Networks, Inc.
- 11/2022-07/2023: Reseach Engineer, Part-time, DATAGRID Inc.
Grants/Awards
- 04/2025-present: Doctoral Course (DC1) Research Fellowship (708/4959=14.3%), Japan Society for the Promotion of Science (JSPS). 「音楽知識を有する記号音楽と言語のマルチモーダルモデルの作成」
- 08/2024: Student Encouragement Award, 第141回音楽情報科学研究発表会 (SIGMUS).
- 08/2023: Encouragement Award (20/139=14.4%), NLP若手の会 (YANS) 第18回シンポジウム.
- 03/2023: Outstanding Award (13/579=2.2%), 言語処理学会第29回年次大会.
Publications
Preprints
-
Tatsuro Inaba, Kentaro Inui, Yusuke Miyao, Yohei Oseki, Benjamin Heinzerling, and Yu Takagi.
"How LLMs Learn: Tracing Internal Representations with Sparse Autoencoders."
[arXiv]
Refereed Papers
-
Go Kamoda, Benjamin Heinzerling, Tatsuro Inaba, Keito Kudo, Keisuke Sakaguchi, and Kentaro Inui.
"Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference."
The 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL 2025 findings)
[arXiv] -
Tatsuro Inaba, Kazuyoshi Yoshii, and Eita Nakamura.
"On the Importance of Time and Pitch Relativity for Transformer-Based Symbolic Music Generation."
The 16th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA 2024)
[paper] -
Tatsuro Inaba, Hirokazu Kiyomaru, Fei Cheng, and Sadao Kurohashi.
"MultiTool-CoT: GPT-3 Can Use Multiple External Tools with Chain of Thought Prompting"
The 61st Annual Meeting of the Association for Computational Linguistics (ACL 2023 main)
[code, arXiv, poster]
Domestic Conference (Japan)
-
稲葉 達郎, 乾 健太郎, 宮尾 祐介, 大関 洋平, Benjamin Heinzerling, 高木 優.
「スパースオートエンコーダーを用いた大規模言語モデルのチェックポイント横断分析」
言語処理学会第31回年次大会．2025.
[paper] -
鴨田 豪, Benjamin Heinzerling, 稲葉 達郎, 工藤 慧音, 坂口 慶祐, 乾 健太郎.
「言語モデルのパラメータから探るDetokenizationメカニズム」
言語処理学会第31回年次大会．2025.
[paper] -
稲葉 達郎, 吉井 和佳, 中村 栄太.
「音楽生成における時間と音高相対性の重要性」
第141回音楽情報科学研究発表会 (SIGMUS)．2024.
[paper] -
稲葉 達郎, 清丸 寛一, Fei Cheng, 黒橋 貞夫.
「大規模言語モデルに基づく複数の外部ツールを利用した推論フレームワーク」
言語処理学会第29回年次大会．2023.
[paper, poster] -
稲葉 達郎, 藤井 拓郎, 小原 涼馬, 柴田 幸輝.
「三言語モデル寄れば文殊の知恵を」
NLP若手の会 (YANS) 第18回シンポジウム, 2023.
[poster]
Talks/Activities
- 06/2025: Talk and tutorials(Pytorch/Transformer/Music Generation) at Kyushu University. [talk slides]
- 03/2025: Talk at Tohoku NLP Group. [slides]
- 08/2024: 論文紹介．第16回最先端NLP勉強会 [slides]
- 09/2023: 学会記事．「大規模言語モデルに基づく複数の外部ツールを利用した推論フレームワーク」の研究．自然言語処理 30巻 3号 p.1100-1104．[paper]
- 08/2023: 論文紹介．第15回最先端NLP勉強会 [slides]