About

My research mainly focuses on the development, application, and analysis of large vision-language models (LVLMs).
I am a third-year Ph.D. student at the Graduate School of Information Sciences, Tohoku University, advised by Prof. Jun Suzuki.

Affiliation

  • Tohoku University, Graduate School of Information Sciences — Ph.D. Student, Fundamental Artificial Intelligence Laboratory
  • University of Cambridge, Department of Computer Science and Technology — Visiting Researcher (Departmental Visiting Student)
  • Turing Inc. — Machine Learning Research Engineer (Intern / Contractor), Fundamental AI Group
Download CV

Research Interests

  • Vision & Language
  • NLP

News

  1. 2026.10

    Our paper “It's Not RoPE That Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention” was accepted to EMNLP 2026.

  2. 2026.09

    Started a research visit at the University of Cambridge (Department of Computer Science and Technology) as a visiting researcher, hosted by Prof. Andreas Vlachos.

  3. 2026.09

    Our book “マルチモーダルモデル入門” (Introduction to Multimodal Models) was published by Gijutsu-Hyoron.

  4. 2026.09

    Our paper “MOMIJI: A Large-Scale Japanese Interleaved Image-Text Dataset” was accepted to the Journal of Natural Language Processing.

  5. 2026.09

    Our paper “Read or Ignore? A Unified Benchmark for Typographic-Attack Robustness and Text Recognition in Vision-Language Models” was accepted to ECCV 2026.

  6. 2026.06

    Our CVPR 2026 paper “Text-Printed Image” received the CVPR Compute Gold Star Award.

Show older news

  1. 2026.03

    Our team won 1st place in the open division and 2nd place overall in the FT-LLM2026 Math Task.

  2. 2026.03

    Four co-authored papers were presented at NLP 2026; two of them received the Young Researcher Encouragement Award.

  3. 2026.01

    Our paper “STRIDE-QA” was presented at AAAI 2026, and our journal paper “Instruction-Following Evaluation of Large Vision-Language Models” was published in New Generation Computing.

  4. 2025.11

    Our paper “Batch-wise Convergent Pre-training” was presented at the First BabyLM Workshop.

  5. 2025.08

    Gave a talk introducing the CVPR 2025 paper “Words or Vision: Do Vision-Language Models Have Blind Faith in Text?” at the 17th Saisentan NLP Study Group (SNLP 2025).

Education

  1. Apr 2024 – Mar 2027 (expected)

    Tohoku University (Ph.D.)

    Ph.D. in Information Science, Fundamental Artificial Intelligence, Department of System Information Sciences, Graduate School of Information Sciences

    Advisor: Prof. Jun Suzuki

  2. Apr 2022 – Mar 2024

    Tohoku University (M.S.)

    M.S. in Information Science, Fundamental Artificial Intelligence, Department of System Information Sciences, Graduate School of Information Sciences

    Advisor: Prof. Jun Suzuki

  3. Apr 2018 – Mar 2022

    Tohoku University (B.E.)

    B.E. in Information Science, Department of Electrical, Information and Physics Engineering, School of Engineering

Experience

  1. Sep 2026 – Present

    Visiting Researcher (Departmental Visiting Student) @ University of Cambridge

    Department of Computer Science and Technology

    Conducting research on Vision & Language and NLP, hosted by Prof. Andreas Vlachos and working closely with Diana Galván Sosa.

  2. Apr 2024 – Present

    Machine Learning Research Engineer Intern / Contractor @ Turing Inc.

    Fundamental AI Group

    Contributed to training large-scale vision-language models (LVLMs), building a large-scale Japanese vision-language pretraining dataset, and constructing a Japanese visual instruction-tuning dataset.

  3. Apr 2023 – Mar 2027 (expected)

    Research Assistant @ Center for Data-driven Science and Artificial Intelligence, Tohoku University

    Supported research activities through experimental setup, documentation, data maintenance, and related project operations.

  4. Mar 2023 – Dec 2023

    Part-time Machine Learning Engineer @ Future Inc.

    Strategic AI Group

    Improved OCR models and developed a model for converting foreign personal names into Japanese readings.

Publications

Underline: myself, *: equal contribution

Journal Papers (refereed)3

  1. JNLP 2026

    MOMIJI: A Large-Scale Japanese Interleaved Image-Text Dataset

    Daiki Shiono, Shingo Yokoi, Shintaro Inuzuka, Tsubasa Takahashi, Jun Suzuki, Yu Yamaguchi.

    Journal of Natural Language Processing, Volume 33, Issue 4 (General Paper; scheduled for publication December 15, 2026). Accepted September 18, 2026.

    acceptance listing train dataset code

  2. NGC 2026

    Instruction-Following Evaluation of Large Vision-Language Models

    Daiki Shiono, Shumpei Miyawaki, Ryota Tanaka, Jun Suzuki.

    New Generation Computing, Volume 44, Article 7, January 2026.

    pdf train dataset

  3. Advanced Robotics 2023

    Aoba_v3 bot: a multimodal chatbot system combining rules and various response generation models

    Shoji Moriya*, Daiki Shiono*, Riki Fujihara, Yosuke Kishinami, Subaru Kimura, Shusaku Sone, Reina Akama, Yuta Matsumoto, Jun Suzuki, Kentaro Inui. (*: equal contribution)

    Advanced Robotics, Volume 37, Issue 21, pp.1392-1405, August 2023.

    pdf

International Conferences (refereed)9

  1. EMNLP 2026

    It's Not RoPE That Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention

    Raito Kiya, Satoki Ohashi, Kosuke Sato, Go Kamoda, Ryosuke Takahashi, Yuji Yamamoto, Daiki Shiono, Keisuke Sakaguchi, Goro Kobayashi.

    Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026), Budapest, Hungary, October 2026.

    pdf code

  2. ECCV 2026

    Read or Ignore? A Unified Benchmark for Typographic-Attack Robustness and Text Recognition in Vision-Language Models

    Futa Waseda, Shojiro Yamabe, Daiki Shiono, Kento Sasaki, Tsubasa Takahashi.

    Proceedings of the European Conference on Computer Vision (ECCV 2026), Malmö, Sweden, September 2026.

    pdf project poster slide train/eval dataset code

  3. CVPR 2026

    Text-Printed Image: Bridging the Image-Text Modality Gap for Text-centric Training of Large Vision-Language Models

    Shojiro Yamabe, Futa Waseda, Daiki Shiono, Tsubasa Takahashi.

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026), Denver, USA, June 2026.

    pdf poster CVPR Compute Gold Star Award

  4. AAAI 2026

    STRIDE-QA: Visual Question Answering Dataset for Spatiotemporal Reasoning in Urban Driving Scenes

    Keishi Ishihara*, Kento Sasaki*, Tsubasa Takahashi, Daiki Shiono, Yu Yamaguchi. (*: equal contribution)

    Proceedings of the AAAI Conference on Artificial Intelligence (AAAI 2026), Singapore, January 2026.

    pdf project slide train dataset eval dataset code

  5. BabyLM 2025

    Batch-wise Convergent Pre-training: Step-by-Step Learning Inspired by Child Language Development

    Ko Yoshida, Daiki Shiono, Kai Sato, Toko Miura, Momoka Furuhashi, Jun Suzuki.

    Proceedings of the First BabyLM Workshop, Association for Computational Linguistics, Suzhou, China, pp.508-524, November 2025.

    pdf poster

  6. COLING 2025

    Evaluating Model Alignment with Human Perception: A Study on Shitsukan in LLMs and LVLMs

    Daiki Shiono, Ana Brassard, Yukiko Ishizuki, Jun Suzuki.

    Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), Abu Dhabi, UAE, January 2025.

    pdf poster eval dataset code

  7. NAACL SRW 2024

    Instruction-Following Evaluation for Large Vision-Language Models

    Daiki Shiono, Shumpei Miyawaki, Ryota Tanaka, Jun Suzuki.

    Non-archival submission for the 2024 Conference of the North American Chapter of the Association for Computational Linguistics Student Research Workshop (NAACL SRW 2024), Mexico City, Mexico, June 2024.

  8. NAACL SRW 2024

    Detecting Response Generation Not Requiring Factual Judgment

    Ryohei Kamei, Daiki Shiono, Reina Akama, Jun Suzuki.

    Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics Student Research Workshop (NAACL SRW 2024), Mexico City, Mexico, June 2024.

    pdf poster

  9. EACL SRW 2024

    Exploring Large Language Models' Understanding of Shitsukan

    Yukiko Ishizuki*, Daiki Shiono*, Ana Brassard, Jun Suzuki. (*: equal contribution)

    Non-archival submission for the 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL SRW 2024), Malta, March 2024.

Domestic Conferences / Symposium / Workshop14

  1. NLP 2026

    MOMIJI: 日本語大規模インターリーブ視覚言語データセット

    Daiki Shiono, Shingo Yokoi, Shintaro Inuzuka, Tsubasa Takahashi, Jun Suzuki, Yu Yamaguchi.

    言語処理学会第32回年次大会 (NLP 2026), Utsunomiya, Tochigi, pp.717-722, March 2026.

    pdf poster train dataset

  2. NLP 2026

    JCuisines: 日本の地域文化の理解にむけた郷土料理視覚言語ベンチマーク

    Kazuki Inomata, Haruto Yoshida, Daiki Shiono, Reina Akama, Keisuke Sakaguchi.

    言語処理学会第32回年次大会 (NLP 2026), Utsunomiya, Tochigi, pp.711-716, March 2026.

    pdf poster

  3. NLP 2026

    注意機構における Attention Sink のバイアス項的解釈

    Satoki Ohashi, Raito Kiya, Kosuke Sato, Ryosuke Takahashi, Go Kamoda, Yuji Yamamoto, Daiki Shiono, Keisuke Sakaguchi, Goro Kobayashi.

    言語処理学会第32回年次大会 (NLP 2026), Utsunomiya, Tochigi, pp.2591-2596, March 2026.

    pdf poster slide Young Researcher Encouragement Award

  4. NLP 2026

    Attention Sink および Massive Activation の発生機序の分解

    Raito Kiya, Satoki Ohashi, Kosuke Sato, Go Kamoda, Ryosuke Takahashi, Yuji Yamamoto, Daiki Shiono, Keisuke Sakaguchi, Goro Kobayashi.

    言語処理学会第32回年次大会 (NLP 2026), Utsunomiya, Tochigi, pp.2579-2584, March 2026.

    pdf poster slide Young Researcher Encouragement Award

  5. FT-LLM 2026

    チーム dentaku FT-LLM 2026 (数学タスク) システム解説

    Keito Kudo, Masatoshi Hidaka, Daiki Shiono, Go Kamoda, Mengyu Ye, Ryosuke Takahashi, Yoichi Aoki, Kazuki Yano, Masaya Taniguchi, Takumi Ito, Keisuke Sakaguchi, Jun Suzuki.

    The Second Workshop on Fine-Tuning and Evaluation of Large Language Models (FT-LLM 2026), Utsunomiya, Tochigi, March 2026.

    pdf poster slide collection code 1st Place in the Open Division, 2nd Place Overall

  6. YANS 2025

    バッチ単位収束型事前学習：子どもの言語発達に着想を得た一歩ずつの学習

    Ko Yoshida, Daiki Shiono, Kai Sato, Momoka Furuhashi, Toko Miura, Jun Suzuki.

    NLP若手の会第20回シンポジウム (YANS 2025), Hamamatsu, Shizuoka, August 2025.

    Encouragement Award

  7. NLP 2025

    Padding vs. Packing: 大規模言語モデルのファインチューニングにおける学習効果の検証

    Daiki Shiono, Ryota Tanaka, Shumpei Miyawaki, Keito Kudo, Jun Suzuki.

    言語処理学会第31回年次大会 (NLP 2025), Nagasaki, pp.1698-1703, March 2025.

    pdf poster

  8. NLP 2025

    大規模視覚言語モデルの質感知覚能力の分析

    Ryosuke Matsuda, Daiki Shiono, Ana Brassard, Jun Suzuki.

    言語処理学会第31回年次大会 (NLP 2025), Nagasaki, pp.2550-2555, March 2025.

    pdf poster

  9. NLP 2025

    日本語バイト符号化マスク言語モデルの開発と分析

    Keito Kudo, Go Kamoda, Daiki Shiono, Jun Suzuki.

    言語処理学会第31回年次大会 (NLP 2025), Nagasaki, pp.3356-3361, March 2025.

    pdf

  10. NLP 2024

    大規模視覚言語モデルに関する指示追従能力の検証

    Daiki Shiono, Shumpei Miyawaki, Ryota Tanaka, Jun Suzuki.

    言語処理学会第30回年次大会 (NLP 2024), Kobe, pp.2057-2062, March 2024.

    pdf poster Committee Special Award

  11. NLP 2024

    事実正誤判定が不要な生成応答の検出に向けたデータセットの収集と分析

    Ryohei Kamei, Daiki Shiono, Reina Akama, Jun Suzuki.

    言語処理学会第30回年次大会 (NLP 2024), Kobe, pp.2125-2130, March 2024.

    pdf slide

  12. YANS 2023

    生成応答に含まれる事実に基づかない情報の自動検出の試み

    Ryohei Kamei, Daiki Shiono, Reina Akama, Jun Suzuki.

    NLP若手の会第18回シンポジウム (YANS 2023), Tokyo, August 2023.

    Encouragement Award CyberAgent Award

  13. NLP 2023

    画像キャプションを利用したIconQAタスクへのアプローチ

    Daiki Shiono, Shumpei Miyawaki, Haruki Nagasawa, Jun Suzuki.

    言語処理学会第29回年次大会 (NLP 2023), Ginowan, Okinawa, pp.1857-1862, March 2023.

    pdf slide

  14. SIG-SLUD 2022

    aoba_v3 bot: 多様な応答生成モデルとルールベースを統合したマルチモーダル雑談対話システム

    Shoji Moriya*, Daiki Shiono*, Yosuke Kishinami, Riki Fujihara, Subaru Kimura, Yuta Matsumoto, Shusaku Sone, Reina Akama, Jun Suzuki, Kentaro Inui. (*: equal contribution)

    人工知能学会 音声・言語理解と対話処理研究会第96回研究会 (JSAI SIG-SLUD 96) 第13回対話システムシンポジウム, Tachikawa, Tokyo, December 2022.

    pdf slide Excellence Award, 2nd Place (Open Track, 5th Dialogue System Live Competition)

Book1

  1. Book 2026

    マルチモーダルモデル入門 (Introduction to Multimodal Models)

    Shuhei Kurita, Seitaro Shinagawa, Rintaro Yanagi, Daiki Shiono, Shinnosuke Takamichi, Daisuke Niizumi, Tomohiko Nakamura, Takuma Yagi, Takayuki Shinohara.

    Gijutsu-Hyoron Co., Ltd., September 28, 2026. ISBN 978-4-297-15836-1.

    Amazon publisher

Projects

  1. Ongoing

    Building a Large-Scale Collection of SFT Datasets Translated into Japanese

    Independent project (FineVision_JA).

    train dataset

  2. Nov 2024

    Sharing Slides Summarizing Insights on LVLMs as of 2024

    Independent project: “Large Vision Language Model (LVLM) に関する最新知見まとめ (Part 1)”

    slide

  3. Sep 2024

    Building SFT Datasets for Japanese LVLMs

    Outcome from my internship at Turing Inc.: The Cauldron JA and Wikipedia Vision JA.

    blog

  4. Jun 2024

    Building a Japanese BERT Model for Long-Sequence Processing

    Collaborated on the development and distributed pretraining of a Japanese BERT model supporting long-sequence inputs, using Megatron-LM on the TSUBAME4.0 supercomputer.

    model

Awards

  1. Mar 2026

    1st Place in the Open Division and 2nd Place Overall, FT-LLM2026 Math Task

  2. Mar 2024

    Committee Special Award, 言語処理学会第30回年次大会 (NLP 2024)

    (26/567 ≈ 4.59%)

  3. Dec 2022

    Excellence Award, 2nd Place, Open Track, 5th Dialogue System Live Competition

Talks

  1. Aug 2025

    Words or Vision: Do Vision-Language Models Have Blind Faith in Text? (CVPR 2025)

    The 17th Saisentan NLP Study Group (SNLP 2025)

    slide

Activities

Research Fellowship / Scholarship

  1. Sep 2024 – Mar 2027 (expected)

    BOOST Program for Next-Generation AI Areas

    ※ BOOST: Broadening Opportunities for Outstanding young researchers and doctoral students in STrategic areas

    Research Fellow (Subsidized Project by the JST (Japan Science and Technology Agency))

  2. Apr 2020 – Mar 2025

    WISE Program for AI Electronics (AIE)

    Research Fellow (Subsidized Project by the Ministry of Education, Culture, Sports, Science and Technology)

  3. Apr 2022 – Mar 2024

    Japan Student Services Organization (JASSO) Scholarship

    Scholarship. Half Repayment Exemption due to outstanding achievements.

Teaching Assistant

  1. Oct 2023 – Feb 2024

    Natural Language Processing

    Graduate-level course, Graduate School of Information Sciences, Tohoku University

  2. Oct 2022 – Dec 2023

    Introduction to Programming for Artificial Intelligence Technology

    General education course, Tohoku University

Skills

  • Python
  • PyTorch
  • Transformers
  • Linux
  • Docker
  • Singularity