Peer-reviewed publications
【査読付き出版物】
Akito Taneguchi, Itsumi Saito, Haruto Yoshida, Jun Suzuki. VisEditBench: A Benchmark for Vector-Format Diagram Editing with Visual Instructions. The Fortieth Annual Conference on Neural Information Processing Systems (NeurIPS 2026) Evaluations and Datasets Track, Dec 2026.
Kamila Szeląg-Levine, Steven Coyne, Tomotaka Kubota, Shinnosuke Nozue, Hong Hai Ngo, Machi Shimmei, Yuichiroh Matsubayashi, Keisuke Sakaguchi. LLM-Assisted Annotation of Japanese Grammatical Patterns and Their Senses for Corpus Construction (APCLC 2026), September 2026.
Kento Haneda, Yosuke Kishinami, Ryo Fujii, Keisuke Sakaguchi, Makoto Morishita. WikiOriginQA: Automated Construction of a QA Dataset with Knowledge Origins for Analyzing Cultural Bias. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP 2026), November 2026.
Yifeng Wang, Zhun Sun, Hakaze Cho, Linlin Wang, Keisuke Sakaguchi. Activation Residual Hessian Quantization for Low-Bit Large Language Model Inference. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP 2026), November 2026.
Subaru Kimura, Ryota Tanaka, Shumpei Miyawaki, Jun Suzuki, Keisuke Sakaguchi. Empirical Analysis of Goal Hijacking in Large Vision-Language Models via Visual Prompt Injection. New Generation Computing.
Momoka Furuhashi, Kouta Nakayama, Noboru Kawai, Takashi Kodama, Saku Sugawara, Kyosuke Takami. Investigating Learner-Aware Design of LLM-Generated Educational Feedback. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP2026). November 2026
Wenzhi Wang, Thiemo Wambsganss, Chenyang Wang, Paul Reisert, Léane Wettstein, Katja Pott, Kenshi Yamaguchi, Machi Shimmei, Roman Rietsche, Surawat Pothong, Naoya Inoue, Kentaro Inui. CAT: A CounterArgument Tutoring System based on Logical Structure Analysis. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations (EMNLP 2026). October 2026.
Muhammad Cendekia Airlangga, Hilal AlQuabeh, Munachiso Samuel Nwadike, Kentaro Inui. Sparse Long Memory and Transition Bottlenecks Shape Positional Recall in Mamba Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Jinghui Zhang, Lang Gao, Ao Li, Mingzhe Li, Ruihong Zeng, Zirui Song, Kentaro Inui, Xiuying Chen. LITERARYBIGFIVE: Author-Personalized Text Generation in a Unified Interpretable Space. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Kareem Elozeiri, Mervat Abassy, Omar Kallas, Fahim Dalvi, Preslav Nakov, Kentaro Inui, Nadir Durrani. Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Sebastian Gerstner, Hilal AlQuabeh, Kentaro Inui, Hinrich Schuetze. Weakening Neurons: An Input-Output Functionality in Transformers with Outsize Influence. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Joe Stacey, Hadas Orgad, Kentaro Inui, Benjamin Heinzerling, Nafise Sadat Moosavi. A Robust Evaluation of Probe Robustness: Lessons for Reliable OOD Uncertainty Quantification. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Tarek Naous, Anagha Savit, Carlos Rafael Catalan, Geyang Guo, Jaehyeok Lee, Kyungdon Lee, Lheane Marie Dizon, Mengyu Ye, Neel Kothari, Sahajpreet Singh, Sarah Masud, Tanish Patwa, Trung Thanh Tran, Zohaib Khan, Alan Ritter, Tanmoy Chakraborty, Yuki Arase, Keisuke Sakaguchi, JinYeong Bak, Wei Xu. Camellia: Benchmarking Cultural Biases in LLMs for Asian Languages. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Momoka Furuhashi, Kouta Nakayama, Takashi Kodama, Saku Sugawara, Kyosuke Takami. How Well Can LLMs Simulate Real Learner Evaluations of Educational Feedback? In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Raito Kiya, Satoki Ohashi, Kosuke Sato, Go Kamoda, Ryosuke Takahashi, Yuji Yamamoto, Daiki Shiono, Keisuke Sakaguchi, Goro Kobayashi. It’s Not RoPE That Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui. Signals Are Present, Predictions Fail: Probing Diagram Representations in Large Vision-Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Keisuke Sakaguchi. GRACE: A Tunable Metric Unifying Accuracy and Efficiency for Robotics Evaluation. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026), October, 2026
Shoji Moriya, Takato Yamazaki, Masaya Ohagi, Junya Takayama, Katsumasa Yoshikawa, Keita Higuchi. AdBot: Effects of Separating Conversational and Advertising Agents on Trust. The 14th International Conference on Human-Agent Interaction (HAI2026), November 2026.
Kento Yoshimura, Kouta Nakayama, Takashi Kodama, Momoka Furuhashi, Kyousuke Takami. Investigating Oddness Annotation Ambiguity in LLM-Generated Stories for Elementary School Kanji Learning. The 34th International Conference on Computers in Education (ICCE2026), November 2026.
Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Machi Shimmei, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. How Well Does AI-Generated Feedback Work? Intrinsic and Extrinsic Evaluation across more than 20,000 EFL Essay Drafts. The 27th International Conference on Artificial Intelligence in Education (AIED 2026 Late Breaking Results), pp. 247-253, June 2026.
Tomomasa Hara, Hiroto Kurita, Masaaki Imaizumi, Kentaro Inui, Sho Yokoi. Why Mean Pooling Works: Quantifying Second-Order Collapse in Text Embeddings. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), pages 47180-47201, July 2026.
Machi Shimmei, Masaki Uto, Yuichiroh Matsubayashi, Kentaro Inui, Aditi Mallavarapu, & Noboru Matsuda. Misconception-Aware Prompting Can Generate Authentic Analytical Multiple-Choice Questions. IEEE Access, 14, 53513–53528 (2026).
古橋萌々香, 舟山弘晃, 松林優一郎, 震明万智, 磯部順子, 石井雄隆, 乾 健太郎. 答案診断グラフを用いた国語記述式問題のための論理構造に基づくフィードバック自動生成. 自然言語処理, Vol.33, No.1, March 2026.
Shinnosuke Nozue*, Yuto Nakano*, Yotaro Watanabe, Meguru Takasaki, Shoji Moriya, Reina Akama and Jun Suzuki. Enhancing Persuasive Dialogue Agents by Synthesizing Cross‑Disciplinary Communication Strategies. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP2025): Industry Track, pages 2287-2312, November 2025. *: Equal Contribution.
Momoka Furuhashi, Kouta Nakayama, Takashi Kodama, and Saku Sugawara. Are Checklists Really Useful for Automatic Evaluation of Generative Tasks? The 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP2025), November 2025.
Ko Yoshida, Daiki Shiono, Kai Sato, Toko Miura, Momoka Furuhashi, and Jun Suzuki. 2025. Batch-wise Convergent Pre-training: Step-by-Step Learning Inspired by Child Language Development. In Proceedings of the First BabyLM Workshop, pages 508–524, Suzhou, China. Association for Computational Linguistics.
Suchun Xie, Shota Sasaki, Hwichan Kim, Yunmeng Li, Reina Akama, Jun Suzuki. Understanding Cross-Lingual Generalization of English-Centric LLMs: The Role of Representation Similarity and Data Exposure. In The 2025 Pacific Rim International Conference on Artificial Intelligence (PRICAI 2025), November 2025.
Wataru Ikeda, Kazuki Yano, Ryosuke Takahashi, Jaesung Lee, Keigo Shibata, and Jun Suzuki. Layer-wise Importance Analysis of Feed-Forward Networks in Transformer-based Language Models. Conference on Language Modeling (COLM 2025), October 2025.
Ikumi Numaya, Shoji Moriya, Shiki Sato, Reina Akama and Jun Suzuki. How Stylistic Similarity Shapes Preferences in Dialogue Dataset with User and Third Party Evaluations. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 192–205, August 2025.
Tetsuro Takahashi, Hirofumi Kikuchi, Jie Yang, Hiroyuki Nishikawa, Masato Komuro, Ryosaku Makino, Shiki Sato, Yuta Sasaki, Shinji Iwata, Asahi Hentona, Takato Yamazaki, Shoji Moriya, Masaya Ohagi, Zhiyang Qi, Takashi Kodama, Akinobu Lee, Takashi Minato, Kurima Sakai, Tomo Funayama, Kotaro Funakoshi, Mayumi Usami, Michimasa Inaba and Ryuichiro Higashinaka. Analyzing Dialogue System Behavior in a Specific Situation Requiring Interpersonal Consideration. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 166–171, August 2025.
Shiki Sato, Shinji Iwata, Asahi Hentona, Yuta Sasaki, Takato Yamazaki, Shoji Moriya, Masaya Ohagi, Hirofumi Kikuchi, Jie Yang, Zhiyang Qi, Takashi Kodama, Akinobu Lee, Masato Komuro, Hiroyuki Nishikawa, Ryosaku Makino, Takashi Minato, Kurima Sakai, Tomo Funayama, Kotaro Funakoshi, Mayumi Usami, Michimasa Inaba, Tetsuro Takahashi and Ryuichiro Higashinaka. Key Challenges in Multimodal Task-Oriented Dialogue Systems: Insights from a Large Competition-Based Dataset. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 449–460, August 2025.
Diana Galvan-Sosa, Gabrielle Gaudeau, Pride Kavumba, Yunmeng Li, Hongyi Gu, Zheng Yuan, Keisuke Sakaguchi, and Paula Buttery. Rubrik’s Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), July 2025.
Masaki Sakata, Benjamin Heinzerling, Sho Yokoi, Takumi Ito & Kentaro Inui. On Entity Identification in Language Models. In Findings of the 63rd Annual Meeting of the Association for Computational Linguistics. (ACL2025 Findings), pp.-, July 2025.
Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Camélia Guerraoui, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. Annotating Errors in English Learners’ Written Language Production: Advancing Automated Written Feedback Systems. The 26th International Conference on Artificial Intelligence in Education (AIED 2025), pp. 292-306, July 2025.
Itsumi Saito, Haruto Yoshida, Keisuke Sakaguchi. Sketch2Diagram: Generating Vector Diagrams from Hand-Drawn Sketches. In The Thirteenth International Conference on Learning Representations (ICLR 2025), April 2025.
Go Kamoda, Heinzerling Benjamin, Tatsuro Inaba, Keito Kudo, Keisuke Sakaguchi, Kentaro Inui. Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference. In Findings of the Association for Computational Linguistics: NAACL 2025. pp.6324-6343, April 2025.
Hirohane Takagi, Shoji Moriya, Takuma Sato, Manabu Nagao, Keita Higuchi. A Framework for Efficient Development and Debugging of Role-Playing Agents with Large Language Models. The 30th Annual ACM Conference on Intelligent User Interfaces (IUI 2025) , pp.70-88, March 2025.
Yunmeng Li, Jun Suzuki, Makoto Morishita, Kaori Abe and Kentaro Inui. MQM-Chat: Multidimensional Quality Metrics for Chat Translation.In Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), pp.3283–3299, January 2025.
Go Kamoda, Akari Asai, Ana Brassard, Keisuke Sakaguchi. Quantifying the Influence of Evaluation Aspects on Long-Form Response Assessment. In Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), pp.8787–8808, January 2025.
Sho Yokoi, Han Bao, Hiroto Kurita, Hidetoshi Shimodaira. Zipfian Whitening. In Advances in Neural Information Processing Systems 37 (NeurIPS 2024), pp.122259–122291, December 2024.
Yoichi Aoki, Keito Kudo, Tatsuki Kuribayashi, Shusaku Sone, Masaya Taniguchi, Keisuke Sakaguchi, Kentaro Inui. First Heuristic Then Rational: Dynamic Use of Heuristics in Language Model Reasoning. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), pp.14255–14271, November 2024
Qin Dai, Benjamin Heinzerling and Kentaro Inui.Representational Analysis of Binding in Language Models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), pp.17468-17493, November 2024
Irfan Robbani, Paul Reisert, Surawat Pothong, Naoya Inoue, Camélia Guerraoui, Wenzhi Wang, Shoichi Naito, Jungmin Choi, Kentaro Inui. Flee the Flaw: Annotating the Underlying Logic of Fallacious Arguments Through Templates and Slot-filling. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024),pp.20524-20540, November 2024.
Keito Kudo*, Hiroyuki Deguchi*, Makoto Morishita*, Ryo Fujii*, Takumi Ito*, Shintaro Ozaki*, Koki Natsumi*, Kai Sato, Kazuki Yano, Ryosuke Takahashi, Subaru Kimura, Tomomasa Hara, Yusuke Sakai, Jun Suzuki (*equal contribution). Document-level Translation with LLM Reranking: Team-J at WMT 2024 General Translation Task. In Proceedings of Ninth Conference on Machine Translation (WMT-2024), pp.210-226, November 2024.
Ana Brassard, Benjamin Heinzerling, Keito Kudo, Keisuke Sakaguchi, Kentaro Inui. ACORN: Aspect-wise Commonsense Reasoning Explanation Evaluation. Conference on Language Modeling (COLM 2024), October 2024.
Shinnosuke Nozue, Yuto Nakano, Shoji Moriya, Tomoki Ariyama, Kazuma Kokuta, Suchun Xie, Kai Sato, Shusaku Sone, Ryohei Kamei, Reina Akama, Yuichiroh Matsubayashi, Keisuke Sakaguchi. A Multimodal Dialogue System to Lead Consensus Building with Emotion-Displaying. The Special Interest Group in discourse and dialogue (SIGDIAL 2024), pp. 669-673, September 2024.
Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui. How Well Do Vision Models Encode Diagram Attributes?. Non-archival submission for ACL Student Research Workshop (ACL-SRW 2024) August 2024.
Kazuki Yano, Takumi Ito, Jun Suzuki. STEP: Staged Parameter-Efficient Pre-training for Large Language Models. Non-archival submission for ACL Student Research Workshop (ACL-SRW 2024) pp.374-384, August 2024.
Kosuke Matsuzaki, Masaya Taniguchi, Kentaro Inui, Keisuke Sakaguchi. J-UniMorph: Japanese Morphological Annotation through the Universal Feature Schema. 21st SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology, pp. 7–19, June 2024.
Subaru Kimura, Ryota Tanaka, Shumpei Miyawaki, Jun Suzuki, Keisuke Sakaguchi. Empirical Analysis of Large Vision Language Models against Goal Hijacking via Visual Prompt Injection. Non-archival submission for NAACL Student Research Workshop (NAACL-SRW 2024), June 2024.
Ryosuke Takahashi, Go Kamoda, Benjamin Heinzerling, Keisuke Sakaguchi, Kentaro Inui. The Curse of Popularity: Popular Entities have Catastrophic Side Effects when Deleting Knowledge from Language Models. Non-archival submission for NAACL Student Research Workshop (NAACL-SRW 2024), 7pages, June 2024.
Naoki Miura, Hiroaki Funayama, Seiya Kikuchi, Yuichiroh Matsubayashi, Yuya Iwase, Kentaro Inui. Japanese-English Sentence Translation Exercises Dataset for Automatic Grading. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL-SRW 2024), pp.266–278, March 2024.
Masahiro Makino, Yuya Asazuma, Shota Sasaki, Jun Suzuki. The Impact of Integration Step on Integrated Gradients.The 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL-SRW 2024), pp.279-289, March 2024.
Peer-reviewed publications
【査読付き出版物】
Akito Taneguchi, Itsumi Saito, Haruto Yoshida, Jun Suzuki. VisEditBench: A Benchmark for Vector-Format Diagram Editing with Visual Instructions. The Fortieth Annual Conference on Neural Information Processing Systems (NeurIPS 2026) Evaluations and Datasets Track, Dec 2026.
Kamila Szeląg-Levine, Steven Coyne, Tomotaka Kubota, Shinnosuke Nozue, Hong Hai Ngo, Machi Shimmei, Yuichiroh Matsubayashi, Keisuke Sakaguchi. LLM-Assisted Annotation of Japanese Grammatical Patterns and Their Senses for Corpus Construction (APCLC 2026), September 2026.
Kento Haneda, Yosuke Kishinami, Ryo Fujii, Keisuke Sakaguchi, Makoto Morishita. WikiOriginQA: Automated Construction of a QA Dataset with Knowledge Origins for Analyzing Cultural Bias. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP 2026), November 2026.
Yifeng Wang, Zhun Sun, Hakaze Cho, Linlin Wang, Keisuke Sakaguchi. Activation Residual Hessian Quantization for Low-Bit Large Language Model Inference. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP 2026), November 2026.
Subaru Kimura, Ryota Tanaka, Shumpei Miyawaki, Jun Suzuki, Keisuke Sakaguchi. Empirical Analysis of Goal Hijacking in Large Vision-Language Models via Visual Prompt Injection. New Generation Computing.
Momoka Furuhashi, Kouta Nakayama, Noboru Kawai, Takashi Kodama, Saku Sugawara, Kyosuke Takami. Investigating Learner-Aware Design of LLM-Generated Educational Feedback. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP2026). November 2026
Wenzhi Wang, Thiemo Wambsganss, Chenyang Wang, Paul Reisert, Léane Wettstein, Katja Pott, Kenshi Yamaguchi, Machi Shimmei, Roman Rietsche, Surawat Pothong, Naoya Inoue, Kentaro Inui. CAT: A CounterArgument Tutoring System based on Logical Structure Analysis. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations (EMNLP 2026). October 2026.
Muhammad Cendekia Airlangga, Hilal AlQuabeh, Munachiso Samuel Nwadike, Kentaro Inui. Sparse Long Memory and Transition Bottlenecks Shape Positional Recall in Mamba Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Jinghui Zhang, Lang Gao, Ao Li, Mingzhe Li, Ruihong Zeng, Zirui Song, Kentaro Inui, Xiuying Chen. LITERARYBIGFIVE: Author-Personalized Text Generation in a Unified Interpretable Space. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Kareem Elozeiri, Mervat Abassy, Omar Kallas, Fahim Dalvi, Preslav Nakov, Kentaro Inui, Nadir Durrani. Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Sebastian Gerstner, Hilal AlQuabeh, Kentaro Inui, Hinrich Schuetze. Weakening Neurons: An Input-Output Functionality in Transformers with Outsize Influence. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Joe Stacey, Hadas Orgad, Kentaro Inui, Benjamin Heinzerling, Nafise Sadat Moosavi. A Robust Evaluation of Probe Robustness: Lessons for Reliable OOD Uncertainty Quantification. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Tarek Naous, Anagha Savit, Carlos Rafael Catalan, Geyang Guo, Jaehyeok Lee, Kyungdon Lee, Lheane Marie Dizon, Mengyu Ye, Neel Kothari, Sahajpreet Singh, Sarah Masud, Tanish Patwa, Trung Thanh Tran, Zohaib Khan, Alan Ritter, Tanmoy Chakraborty, Yuki Arase, Keisuke Sakaguchi, JinYeong Bak, Wei Xu. Camellia: Benchmarking Cultural Biases in LLMs for Asian Languages. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Momoka Furuhashi, Kouta Nakayama, Takashi Kodama, Saku Sugawara, Kyosuke Takami. How Well Can LLMs Simulate Real Learner Evaluations of Educational Feedback? In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Raito Kiya, Satoki Ohashi, Kosuke Sato, Go Kamoda, Ryosuke Takahashi, Yuji Yamamoto, Daiki Shiono, Keisuke Sakaguchi, Goro Kobayashi. It’s Not RoPE That Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui. Signals Are Present, Predictions Fail: Probing Diagram Representations in Large Vision-Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Keisuke Sakaguchi. GRACE: A Tunable Metric Unifying Accuracy and Efficiency for Robotics Evaluation. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026), October, 2026
Shoji Moriya, Takato Yamazaki, Masaya Ohagi, Junya Takayama, Katsumasa Yoshikawa, Keita Higuchi. AdBot: Effects of Separating Conversational and Advertising Agents on Trust. The 14th International Conference on Human-Agent Interaction (HAI2026), November 2026.
Kento Yoshimura, Kouta Nakayama, Takashi Kodama, Momoka Furuhashi, Kyousuke Takami. Investigating Oddness Annotation Ambiguity in LLM-Generated Stories for Elementary School Kanji Learning. The 34th International Conference on Computers in Education (ICCE2026), November 2026.
Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Machi Shimmei, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. How Well Does AI-Generated Feedback Work? Intrinsic and Extrinsic Evaluation across more than 20,000 EFL Essay Drafts. The 27th International Conference on Artificial Intelligence in Education (AIED 2026 Late Breaking Results), pp. 247-253, June 2026.
Tomomasa Hara, Hiroto Kurita, Masaaki Imaizumi, Kentaro Inui, Sho Yokoi. Why Mean Pooling Works: Quantifying Second-Order Collapse in Text Embeddings. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), pages 47180-47201, July 2026.
Machi Shimmei, Masaki Uto, Yuichiroh Matsubayashi, Kentaro Inui, Aditi Mallavarapu, & Noboru Matsuda. Misconception-Aware Prompting Can Generate Authentic Analytical Multiple-Choice Questions. IEEE Access, 14, 53513–53528 (2026).
古橋萌々香, 舟山弘晃, 松林優一郎, 震明万智, 磯部順子, 石井雄隆, 乾 健太郎. 答案診断グラフを用いた国語記述式問題のための論理構造に基づくフィードバック自動生成. 自然言語処理, Vol.33, No.1, March 2026.
Shinnosuke Nozue*, Yuto Nakano*, Yotaro Watanabe, Meguru Takasaki, Shoji Moriya, Reina Akama and Jun Suzuki. Enhancing Persuasive Dialogue Agents by Synthesizing Cross‑Disciplinary Communication Strategies. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP2025): Industry Track, pages 2287-2312, November 2025. *: Equal Contribution.
Momoka Furuhashi, Kouta Nakayama, Takashi Kodama, and Saku Sugawara. Are Checklists Really Useful for Automatic Evaluation of Generative Tasks? The 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP2025), November 2025.
Ko Yoshida, Daiki Shiono, Kai Sato, Toko Miura, Momoka Furuhashi, and Jun Suzuki. 2025. Batch-wise Convergent Pre-training: Step-by-Step Learning Inspired by Child Language Development. In Proceedings of the First BabyLM Workshop, pages 508–524, Suzhou, China. Association for Computational Linguistics.
Suchun Xie, Shota Sasaki, Hwichan Kim, Yunmeng Li, Reina Akama, Jun Suzuki. Understanding Cross-Lingual Generalization of English-Centric LLMs: The Role of Representation Similarity and Data Exposure. In The 2025 Pacific Rim International Conference on Artificial Intelligence (PRICAI 2025), November 2025.
Wataru Ikeda, Kazuki Yano, Ryosuke Takahashi, Jaesung Lee, Keigo Shibata, and Jun Suzuki. Layer-wise Importance Analysis of Feed-Forward Networks in Transformer-based Language Models. Conference on Language Modeling (COLM 2025), October 2025.
Ikumi Numaya, Shoji Moriya, Shiki Sato, Reina Akama and Jun Suzuki. How Stylistic Similarity Shapes Preferences in Dialogue Dataset with User and Third Party Evaluations. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 192–205, August 2025.
Tetsuro Takahashi, Hirofumi Kikuchi, Jie Yang, Hiroyuki Nishikawa, Masato Komuro, Ryosaku Makino, Shiki Sato, Yuta Sasaki, Shinji Iwata, Asahi Hentona, Takato Yamazaki, Shoji Moriya, Masaya Ohagi, Zhiyang Qi, Takashi Kodama, Akinobu Lee, Takashi Minato, Kurima Sakai, Tomo Funayama, Kotaro Funakoshi, Mayumi Usami, Michimasa Inaba and Ryuichiro Higashinaka. Analyzing Dialogue System Behavior in a Specific Situation Requiring Interpersonal Consideration. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 166–171, August 2025.
Shiki Sato, Shinji Iwata, Asahi Hentona, Yuta Sasaki, Takato Yamazaki, Shoji Moriya, Masaya Ohagi, Hirofumi Kikuchi, Jie Yang, Zhiyang Qi, Takashi Kodama, Akinobu Lee, Masato Komuro, Hiroyuki Nishikawa, Ryosaku Makino, Takashi Minato, Kurima Sakai, Tomo Funayama, Kotaro Funakoshi, Mayumi Usami, Michimasa Inaba, Tetsuro Takahashi and Ryuichiro Higashinaka. Key Challenges in Multimodal Task-Oriented Dialogue Systems: Insights from a Large Competition-Based Dataset. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 449–460, August 2025.
Diana Galvan-Sosa, Gabrielle Gaudeau, Pride Kavumba, Yunmeng Li, Hongyi Gu, Zheng Yuan, Keisuke Sakaguchi, and Paula Buttery. Rubrik’s Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), July 2025.
Masaki Sakata, Benjamin Heinzerling, Sho Yokoi, Takumi Ito & Kentaro Inui. On Entity Identification in Language Models. In Findings of the 63rd Annual Meeting of the Association for Computational Linguistics. (ACL2025 Findings), pp.-, July 2025.
Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Camélia Guerraoui, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. Annotating Errors in English Learners’ Written Language Production: Advancing Automated Written Feedback Systems. The 26th International Conference on Artificial Intelligence in Education (AIED 2025), pp. 292-306, July 2025.
Itsumi Saito, Haruto Yoshida, Keisuke Sakaguchi. Sketch2Diagram: Generating Vector Diagrams from Hand-Drawn Sketches. In The Thirteenth International Conference on Learning Representations (ICLR 2025), April 2025.
Go Kamoda, Heinzerling Benjamin, Tatsuro Inaba, Keito Kudo, Keisuke Sakaguchi, Kentaro Inui. Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference. In Findings of the Association for Computational Linguistics: NAACL 2025. pp.6324-6343, April 2025.
Hirohane Takagi, Shoji Moriya, Takuma Sato, Manabu Nagao, Keita Higuchi. A Framework for Efficient Development and Debugging of Role-Playing Agents with Large Language Models. The 30th Annual ACM Conference on Intelligent User Interfaces (IUI 2025) , pp.70-88, March 2025.
Yunmeng Li, Jun Suzuki, Makoto Morishita, Kaori Abe and Kentaro Inui. MQM-Chat: Multidimensional Quality Metrics for Chat Translation.In Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), pp.3283–3299, January 2025.
Go Kamoda, Akari Asai, Ana Brassard, Keisuke Sakaguchi. Quantifying the Influence of Evaluation Aspects on Long-Form Response Assessment. In Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), pp.8787–8808, January 2025.
Sho Yokoi, Han Bao, Hiroto Kurita, Hidetoshi Shimodaira. Zipfian Whitening. In Advances in Neural Information Processing Systems 37 (NeurIPS 2024), pp.122259–122291, December 2024.
Yoichi Aoki, Keito Kudo, Tatsuki Kuribayashi, Shusaku Sone, Masaya Taniguchi, Keisuke Sakaguchi, Kentaro Inui. First Heuristic Then Rational: Dynamic Use of Heuristics in Language Model Reasoning. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), pp.14255–14271, November 2024
Qin Dai, Benjamin Heinzerling and Kentaro Inui.Representational Analysis of Binding in Language Models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), pp.17468-17493, November 2024
Irfan Robbani, Paul Reisert, Surawat Pothong, Naoya Inoue, Camélia Guerraoui, Wenzhi Wang, Shoichi Naito, Jungmin Choi, Kentaro Inui. Flee the Flaw: Annotating the Underlying Logic of Fallacious Arguments Through Templates and Slot-filling. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024),pp.20524-20540, November 2024.
Keito Kudo*, Hiroyuki Deguchi*, Makoto Morishita*, Ryo Fujii*, Takumi Ito*, Shintaro Ozaki*, Koki Natsumi*, Kai Sato, Kazuki Yano, Ryosuke Takahashi, Subaru Kimura, Tomomasa Hara, Yusuke Sakai, Jun Suzuki (*equal contribution). Document-level Translation with LLM Reranking: Team-J at WMT 2024 General Translation Task. In Proceedings of Ninth Conference on Machine Translation (WMT-2024), pp.210-226, November 2024.
Ana Brassard, Benjamin Heinzerling, Keito Kudo, Keisuke Sakaguchi, Kentaro Inui. ACORN: Aspect-wise Commonsense Reasoning Explanation Evaluation. Conference on Language Modeling (COLM 2024), October 2024.
Shinnosuke Nozue, Yuto Nakano, Shoji Moriya, Tomoki Ariyama, Kazuma Kokuta, Suchun Xie, Kai Sato, Shusaku Sone, Ryohei Kamei, Reina Akama, Yuichiroh Matsubayashi, Keisuke Sakaguchi. A Multimodal Dialogue System to Lead Consensus Building with Emotion-Displaying. The Special Interest Group in discourse and dialogue (SIGDIAL 2024), pp. 669-673, September 2024.
Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui. How Well Do Vision Models Encode Diagram Attributes?. Non-archival submission for ACL Student Research Workshop (ACL-SRW 2024) August 2024.
Kazuki Yano, Takumi Ito, Jun Suzuki. STEP: Staged Parameter-Efficient Pre-training for Large Language Models. Non-archival submission for ACL Student Research Workshop (ACL-SRW 2024) pp.374-384, August 2024.
Kosuke Matsuzaki, Masaya Taniguchi, Kentaro Inui, Keisuke Sakaguchi. J-UniMorph: Japanese Morphological Annotation through the Universal Feature Schema. 21st SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology, pp. 7–19, June 2024.
Subaru Kimura, Ryota Tanaka, Shumpei Miyawaki, Jun Suzuki, Keisuke Sakaguchi. Empirical Analysis of Large Vision Language Models against Goal Hijacking via Visual Prompt Injection. Non-archival submission for NAACL Student Research Workshop (NAACL-SRW 2024), June 2024.
Ryosuke Takahashi, Go Kamoda, Benjamin Heinzerling, Keisuke Sakaguchi, Kentaro Inui. The Curse of Popularity: Popular Entities have Catastrophic Side Effects when Deleting Knowledge from Language Models. Non-archival submission for NAACL Student Research Workshop (NAACL-SRW 2024), 7pages, June 2024.
Naoki Miura, Hiroaki Funayama, Seiya Kikuchi, Yuichiroh Matsubayashi, Yuya Iwase, Kentaro Inui. Japanese-English Sentence Translation Exercises Dataset for Automatic Grading. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL-SRW 2024), pp.266–278, March 2024.
Masahiro Makino, Yuya Asazuma, Shota Sasaki, Jun Suzuki. The Impact of Integration Step on Integrated Gradients.The 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL-SRW 2024), pp.279-289, March 2024.
Peer-reviewed publications
【査読付き出版物】
Akito Taneguchi, Itsumi Saito, Haruto Yoshida, Jun Suzuki. VisEditBench: A Benchmark for Vector-Format Diagram Editing with Visual Instructions. The Fortieth Annual Conference on Neural Information Processing Systems (NeurIPS 2026) Evaluations and Datasets Track, Dec 2026.
Kamila Szeląg-Levine, Steven Coyne, Tomotaka Kubota, Shinnosuke Nozue, Hong Hai Ngo, Machi Shimmei, Yuichiroh Matsubayashi, Keisuke Sakaguchi. LLM-Assisted Annotation of Japanese Grammatical Patterns and Their Senses for Corpus Construction (APCLC 2026), September 2026.
Kento Haneda, Yosuke Kishinami, Ryo Fujii, Keisuke Sakaguchi, Makoto Morishita. WikiOriginQA: Automated Construction of a QA Dataset with Knowledge Origins for Analyzing Cultural Bias. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP 2026), November 2026.
Yifeng Wang, Zhun Sun, Hakaze Cho, Linlin Wang, Keisuke Sakaguchi. Activation Residual Hessian Quantization for Low-Bit Large Language Model Inference. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP 2026), November 2026.
Subaru Kimura, Ryota Tanaka, Shumpei Miyawaki, Jun Suzuki, Keisuke Sakaguchi. Empirical Analysis of Goal Hijacking in Large Vision-Language Models via Visual Prompt Injection. New Generation Computing.
Momoka Furuhashi, Kouta Nakayama, Noboru Kawai, Takashi Kodama, Saku Sugawara, Kyosuke Takami. Investigating Learner-Aware Design of LLM-Generated Educational Feedback. The 5th Asia-Pacific Chapter of the Association for Computational Linguistics & the 15th International Joint Conference on Natural Language Processing (AACL-IJCNLP2026). November 2026
Wenzhi Wang, Thiemo Wambsganss, Chenyang Wang, Paul Reisert, Léane Wettstein, Katja Pott, Kenshi Yamaguchi, Machi Shimmei, Roman Rietsche, Surawat Pothong, Naoya Inoue, Kentaro Inui. CAT: A CounterArgument Tutoring System based on Logical Structure Analysis. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations (EMNLP 2026). October 2026.
Muhammad Cendekia Airlangga, Hilal AlQuabeh, Munachiso Samuel Nwadike, Kentaro Inui. Sparse Long Memory and Transition Bottlenecks Shape Positional Recall in Mamba Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Jinghui Zhang, Lang Gao, Ao Li, Mingzhe Li, Ruihong Zeng, Zirui Song, Kentaro Inui, Xiuying Chen. LITERARYBIGFIVE: Author-Personalized Text Generation in a Unified Interpretable Space. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Kareem Elozeiri, Mervat Abassy, Omar Kallas, Fahim Dalvi, Preslav Nakov, Kentaro Inui, Nadir Durrani. Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Sebastian Gerstner, Hilal AlQuabeh, Kentaro Inui, Hinrich Schuetze. Weakening Neurons: An Input-Output Functionality in Transformers with Outsize Influence. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Joe Stacey, Hadas Orgad, Kentaro Inui, Benjamin Heinzerling, Nafise Sadat Moosavi. A Robust Evaluation of Probe Robustness: Lessons for Reliable OOD Uncertainty Quantification. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Tarek Naous, Anagha Savit, Carlos Rafael Catalan, Geyang Guo, Jaehyeok Lee, Kyungdon Lee, Lheane Marie Dizon, Mengyu Ye, Neel Kothari, Sahajpreet Singh, Sarah Masud, Tanish Patwa, Trung Thanh Tran, Zohaib Khan, Alan Ritter, Tanmoy Chakraborty, Yuki Arase, Keisuke Sakaguchi, JinYeong Bak, Wei Xu. Camellia: Benchmarking Cultural Biases in LLMs for Asian Languages. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Momoka Furuhashi, Kouta Nakayama, Takashi Kodama, Saku Sugawara, Kyosuke Takami. How Well Can LLMs Simulate Real Learner Evaluations of Educational Feedback? In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Raito Kiya, Satoki Ohashi, Kosuke Sato, Go Kamoda, Ryosuke Takahashi, Yuji Yamamoto, Daiki Shiono, Keisuke Sakaguchi, Goro Kobayashi. It’s Not RoPE That Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui. Signals Are Present, Predictions Fail: Probing Diagram Representations in Large Vision-Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026). October 2026.
Keisuke Sakaguchi. GRACE: A Tunable Metric Unifying Accuracy and Efficiency for Robotics Evaluation. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026), October, 2026
Shoji Moriya, Takato Yamazaki, Masaya Ohagi, Junya Takayama, Katsumasa Yoshikawa, Keita Higuchi. AdBot: Effects of Separating Conversational and Advertising Agents on Trust. The 14th International Conference on Human-Agent Interaction (HAI2026), November 2026.
Kento Yoshimura, Kouta Nakayama, Takashi Kodama, Momoka Furuhashi, Kyousuke Takami. Investigating Oddness Annotation Ambiguity in LLM-Generated Stories for Elementary School Kanji Learning. The 34th International Conference on Computers in Education (ICCE2026), November 2026.
Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Machi Shimmei, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. How Well Does AI-Generated Feedback Work? Intrinsic and Extrinsic Evaluation across more than 20,000 EFL Essay Drafts. The 27th International Conference on Artificial Intelligence in Education (AIED 2026 Late Breaking Results), pp. 247-253, June 2026.
Tomomasa Hara, Hiroto Kurita, Masaaki Imaizumi, Kentaro Inui, Sho Yokoi. Why Mean Pooling Works: Quantifying Second-Order Collapse in Text Embeddings. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), pages 47180-47201, July 2026.
Machi Shimmei, Masaki Uto, Yuichiroh Matsubayashi, Kentaro Inui, Aditi Mallavarapu, & Noboru Matsuda. Misconception-Aware Prompting Can Generate Authentic Analytical Multiple-Choice Questions. IEEE Access, 14, 53513–53528 (2026).
古橋萌々香, 舟山弘晃, 松林優一郎, 震明万智, 磯部順子, 石井雄隆, 乾 健太郎. 答案診断グラフを用いた国語記述式問題のための論理構造に基づくフィードバック自動生成. 自然言語処理, Vol.33, No.1, March 2026.
Shinnosuke Nozue*, Yuto Nakano*, Yotaro Watanabe, Meguru Takasaki, Shoji Moriya, Reina Akama and Jun Suzuki. Enhancing Persuasive Dialogue Agents by Synthesizing Cross‑Disciplinary Communication Strategies. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP2025): Industry Track, pages 2287-2312, November 2025. *: Equal Contribution.
Momoka Furuhashi, Kouta Nakayama, Takashi Kodama, and Saku Sugawara. Are Checklists Really Useful for Automatic Evaluation of Generative Tasks? The 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP2025), November 2025.
Ko Yoshida, Daiki Shiono, Kai Sato, Toko Miura, Momoka Furuhashi, and Jun Suzuki. 2025. Batch-wise Convergent Pre-training: Step-by-Step Learning Inspired by Child Language Development. In Proceedings of the First BabyLM Workshop, pages 508–524, Suzhou, China. Association for Computational Linguistics.
Suchun Xie, Shota Sasaki, Hwichan Kim, Yunmeng Li, Reina Akama, Jun Suzuki. Understanding Cross-Lingual Generalization of English-Centric LLMs: The Role of Representation Similarity and Data Exposure. In The 2025 Pacific Rim International Conference on Artificial Intelligence (PRICAI 2025), November 2025.
Wataru Ikeda, Kazuki Yano, Ryosuke Takahashi, Jaesung Lee, Keigo Shibata, and Jun Suzuki. Layer-wise Importance Analysis of Feed-Forward Networks in Transformer-based Language Models. Conference on Language Modeling (COLM 2025), October 2025.
Ikumi Numaya, Shoji Moriya, Shiki Sato, Reina Akama and Jun Suzuki. How Stylistic Similarity Shapes Preferences in Dialogue Dataset with User and Third Party Evaluations. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 192–205, August 2025.
Tetsuro Takahashi, Hirofumi Kikuchi, Jie Yang, Hiroyuki Nishikawa, Masato Komuro, Ryosaku Makino, Shiki Sato, Yuta Sasaki, Shinji Iwata, Asahi Hentona, Takato Yamazaki, Shoji Moriya, Masaya Ohagi, Zhiyang Qi, Takashi Kodama, Akinobu Lee, Takashi Minato, Kurima Sakai, Tomo Funayama, Kotaro Funakoshi, Mayumi Usami, Michimasa Inaba and Ryuichiro Higashinaka. Analyzing Dialogue System Behavior in a Specific Situation Requiring Interpersonal Consideration. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 166–171, August 2025.
Shiki Sato, Shinji Iwata, Asahi Hentona, Yuta Sasaki, Takato Yamazaki, Shoji Moriya, Masaya Ohagi, Hirofumi Kikuchi, Jie Yang, Zhiyang Qi, Takashi Kodama, Akinobu Lee, Masato Komuro, Hiroyuki Nishikawa, Ryosaku Makino, Takashi Minato, Kurima Sakai, Tomo Funayama, Kotaro Funakoshi, Mayumi Usami, Michimasa Inaba, Tetsuro Takahashi and Ryuichiro Higashinaka. Key Challenges in Multimodal Task-Oriented Dialogue Systems: Insights from a Large Competition-Based Dataset. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL 2025), pages 449–460, August 2025.
Diana Galvan-Sosa, Gabrielle Gaudeau, Pride Kavumba, Yunmeng Li, Hongyi Gu, Zheng Yuan, Keisuke Sakaguchi, and Paula Buttery. Rubrik’s Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), July 2025.
Masaki Sakata, Benjamin Heinzerling, Sho Yokoi, Takumi Ito & Kentaro Inui. On Entity Identification in Language Models. In Findings of the 63rd Annual Meeting of the Association for Computational Linguistics. (ACL2025 Findings), pp.-, July 2025.
Steven Coyne, Diana Galvan-Sosa, Ryan Spring, Camélia Guerraoui, Michael Zock, Keisuke Sakaguchi, and Kentaro Inui. Annotating Errors in English Learners’ Written Language Production: Advancing Automated Written Feedback Systems. The 26th International Conference on Artificial Intelligence in Education (AIED 2025), pp. 292-306, July 2025.
Itsumi Saito, Haruto Yoshida, Keisuke Sakaguchi. Sketch2Diagram: Generating Vector Diagrams from Hand-Drawn Sketches. In The Thirteenth International Conference on Learning Representations (ICLR 2025), April 2025.
Go Kamoda, Heinzerling Benjamin, Tatsuro Inaba, Keito Kudo, Keisuke Sakaguchi, Kentaro Inui. Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference. In Findings of the Association for Computational Linguistics: NAACL 2025. pp.6324-6343, April 2025.
Hirohane Takagi, Shoji Moriya, Takuma Sato, Manabu Nagao, Keita Higuchi. A Framework for Efficient Development and Debugging of Role-Playing Agents with Large Language Models. The 30th Annual ACM Conference on Intelligent User Interfaces (IUI 2025) , pp.70-88, March 2025.
Yunmeng Li, Jun Suzuki, Makoto Morishita, Kaori Abe and Kentaro Inui. MQM-Chat: Multidimensional Quality Metrics for Chat Translation.In Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), pp.3283–3299, January 2025.
Go Kamoda, Akari Asai, Ana Brassard, Keisuke Sakaguchi. Quantifying the Influence of Evaluation Aspects on Long-Form Response Assessment. In Proceedings of the 31st International Conference on Computational Linguistics (COLING 2025), pp.8787–8808, January 2025.
Sho Yokoi, Han Bao, Hiroto Kurita, Hidetoshi Shimodaira. Zipfian Whitening. In Advances in Neural Information Processing Systems 37 (NeurIPS 2024), pp.122259–122291, December 2024.
Yoichi Aoki, Keito Kudo, Tatsuki Kuribayashi, Shusaku Sone, Masaya Taniguchi, Keisuke Sakaguchi, Kentaro Inui. First Heuristic Then Rational: Dynamic Use of Heuristics in Language Model Reasoning. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), pp.14255–14271, November 2024
Qin Dai, Benjamin Heinzerling and Kentaro Inui.Representational Analysis of Binding in Language Models. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024), pp.17468-17493, November 2024
Irfan Robbani, Paul Reisert, Surawat Pothong, Naoya Inoue, Camélia Guerraoui, Wenzhi Wang, Shoichi Naito, Jungmin Choi, Kentaro Inui. Flee the Flaw: Annotating the Underlying Logic of Fallacious Arguments Through Templates and Slot-filling. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024),pp.20524-20540, November 2024.
Keito Kudo*, Hiroyuki Deguchi*, Makoto Morishita*, Ryo Fujii*, Takumi Ito*, Shintaro Ozaki*, Koki Natsumi*, Kai Sato, Kazuki Yano, Ryosuke Takahashi, Subaru Kimura, Tomomasa Hara, Yusuke Sakai, Jun Suzuki (*equal contribution). Document-level Translation with LLM Reranking: Team-J at WMT 2024 General Translation Task. In Proceedings of Ninth Conference on Machine Translation (WMT-2024), pp.210-226, November 2024.
Ana Brassard, Benjamin Heinzerling, Keito Kudo, Keisuke Sakaguchi, Kentaro Inui. ACORN: Aspect-wise Commonsense Reasoning Explanation Evaluation. Conference on Language Modeling (COLM 2024), October 2024.
Shinnosuke Nozue, Yuto Nakano, Shoji Moriya, Tomoki Ariyama, Kazuma Kokuta, Suchun Xie, Kai Sato, Shusaku Sone, Ryohei Kamei, Reina Akama, Yuichiroh Matsubayashi, Keisuke Sakaguchi. A Multimodal Dialogue System to Lead Consensus Building with Emotion-Displaying. The Special Interest Group in discourse and dialogue (SIGDIAL 2024), pp. 669-673, September 2024.
Haruto Yoshida, Keito Kudo, Yoichi Aoki, Ryota Tanaka, Itsumi Saito, Keisuke Sakaguchi, Kentaro Inui. How Well Do Vision Models Encode Diagram Attributes?. Non-archival submission for ACL Student Research Workshop (ACL-SRW 2024) August 2024.
Kazuki Yano, Takumi Ito, Jun Suzuki. STEP: Staged Parameter-Efficient Pre-training for Large Language Models. Non-archival submission for ACL Student Research Workshop (ACL-SRW 2024) pp.374-384, August 2024.
Kosuke Matsuzaki, Masaya Taniguchi, Kentaro Inui, Keisuke Sakaguchi. J-UniMorph: Japanese Morphological Annotation through the Universal Feature Schema. 21st SIGMORPHON Workshop on Computational Research in Phonetics, Phonology, and Morphology, pp. 7–19, June 2024.
Subaru Kimura, Ryota Tanaka, Shumpei Miyawaki, Jun Suzuki, Keisuke Sakaguchi. Empirical Analysis of Large Vision Language Models against Goal Hijacking via Visual Prompt Injection. Non-archival submission for NAACL Student Research Workshop (NAACL-SRW 2024), June 2024.
Ryosuke Takahashi, Go Kamoda, Benjamin Heinzerling, Keisuke Sakaguchi, Kentaro Inui. The Curse of Popularity: Popular Entities have Catastrophic Side Effects when Deleting Knowledge from Language Models. Non-archival submission for NAACL Student Research Workshop (NAACL-SRW 2024), 7pages, June 2024.
Naoki Miura, Hiroaki Funayama, Seiya Kikuchi, Yuichiroh Matsubayashi, Yuya Iwase, Kentaro Inui. Japanese-English Sentence Translation Exercises Dataset for Automatic Grading. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL-SRW 2024), pp.266–278, March 2024.
Masahiro Makino, Yuya Asazuma, Shota Sasaki, Jun Suzuki. The Impact of Integration Step on Integrated Gradients.The 18th Conference of the European Chapter of the Association for Computational Linguistics Student Research Workshop (EACL-SRW 2024), pp.279-289, March 2024.
Other Publications and Presentations
【その他の出版物および発表】
吉村 賢人, 高見 享佑, 古橋 萌々香, 中山 功太 , 児玉 貴志. 小学校漢字学習におけるLLM生成物語の「違和感」の類型化とLLM間の比較. 第21回YANSシンポジウム (YANS2026), August 2026.
松﨑 孝介, 谷口 雅弥. 「待つ」は一瞬か数時間か：LLMによるイベント持続時間の評価. 第21回YANSシンポジウム (YANS2026), August 2026.
工藤 慧音, 山口健史, 坂口 慶祐. nanochat-jp: 小規模日本語特化言語モデル統合研究基盤. 第21回YANSシンポジウム (YANS2026), August 2026.
佐々木 睦史, 葉 夢宇, 趙 羽風, 工藤 慧音, 坂口 慶祐, 鈴木 潤. 一様拡散言語モデルの能力獲得過程. 第21回YANSシンポジウム (YANS2026), August 2026.
吉田 遥音, 工藤 慧音, 松田 陵佑, 清水 伸幸, 鈴木 潤. 長尺動画生成タスクにおけるプロンプトと動画間のシーン対応を考慮した自動評価. 第21回YANSシンポジウム (YANS2026), August 2026.
羽根田 賢和, 大竹 孝樹, 塙 一晃, 樋口 泰地, 山口 学. 社会的関心の尺度として Wikipedia 閲覧数を利用した Web メディアの記事人気予測. 2026年度人工知能学会全国大会 (第40回), June 2026.
渡邉 天照, 羽根田 賢和, 坂口 慶祐. TRPGゲームマスターAIの迎合性がプレイヤー体験に及ぼす影響. 2026年度人工知能学会全国大会 (第40回), June 2026.
松﨑 孝介, 谷口 雅弥, 坂口 慶祐. 日本語方言の音声認識における学習データの規模と特性. 言語処理学会第32回年次大会 (NLP2026), March 2026.
Hong Hai Ngo, Yunmeng Li, Steven Coyne, Kamila Szelag-Levine, 坂口 慶祐. Wav2Karaoke: A Pipeline to Align Notes and Lyrics from a Song. 言語処理学会第32回年次大会 (NLP2026), March 2026.
野末 慎之介, 松林 優一郎, 藤井 諒, 岸波 洋介, 森下 睦, 坂口 慶祐. 項の復元は必要か？日英機械翻訳における省略された項の扱いの分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
佐々木 睦史, 中山 功太, 宮尾 祐介, 大関 洋平, 磯沼 大. 排他的逆学習. 言語処理学会第32回年次大会 (NLP2026), March 2026.
澤田 紘希, 佐々木 睦史, 坂口 慶祐, Benjamin Heinzerling. 言語モデルにおける共参照解決を担う注意ヘッドの分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
神林 励旺, 稲葉 達郎, Benjamin Heinzerling, 坂口 慶祐, 乾 健太郎, 磯沼 大. 画像生成学習が視覚理解能力へ与える影響の分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
原知正, 栗田宙人, 今泉允聡, 乾健太郎, 横井祥. なぜ平均プーリングはうまく動くのか？テキスト埋め込みの二次統計量の崩壊の定量化. 言語処理学会第32回年次大会 (NLP2026), March 2026.
木谷頼斗, 大橋諭貴, 佐藤宏亮, 鴨田豪, 高橋良允, 山本悠士, 塩野大輝, 坂口慶祐, 小林悟郎. Attention Sink および Massive Activation の発生機序の分解. 言語処理学会第32回年次大会 (NLP2026), March 2026.
大橋 諭貴, 木谷 頼斗, 佐藤 宏亮, 高橋 良允, 鴨田 豪, 山本 悠士, 塩野 大輝, 坂口 慶祐, 小林 悟郎. 注意機構における Attention Sink のバイアス項的解釈. 言語処理学会第32回年次大会 (NLP2026), March 2026
猪股和樹, 吉田遥音, 塩野大輝, 赤間怜奈, 坂口慶祐. 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. 言語処理学会第32回年次大会 (NLP2026), March 2026.
高橋良允, 稲葉達郎, 乾健太郎, Benjamin Heinzerling. TopK Language Models. 言語処理学会第32回年次大会 (NLP2026), March 2026.
坂田将樹, Benjamin Heinzerling, 伊藤拓海, 横井祥, 乾健太郎. 言語モデルに線形に内在する階層的概念. 言語処理学会第32回年次大会 (NLP2026), March 2026.
青木洋一, 村上聡一朗, 本多右京, 加藤明彦. 広告文生成における多様性と広告品質の関係. 言語処理学会第32回年次大会 (NLP2026), March 2026.
Other Publications and Presentations
【その他の出版物および発表】
吉村 賢人, 高見 享佑, 古橋 萌々香, 中山 功太 , 児玉 貴志. 小学校漢字学習におけるLLM生成物語の「違和感」の類型化とLLM間の比較. 第21回YANSシンポジウム (YANS2026), August 2026.
松﨑 孝介, 谷口 雅弥. 「待つ」は一瞬か数時間か：LLMによるイベント持続時間の評価. 第21回YANSシンポジウム (YANS2026), August 2026.
工藤 慧音, 山口健史, 坂口 慶祐. nanochat-jp: 小規模日本語特化言語モデル統合研究基盤. 第21回YANSシンポジウム (YANS2026), August 2026.
佐々木 睦史, 葉 夢宇, 趙 羽風, 工藤 慧音, 坂口 慶祐, 鈴木 潤. 一様拡散言語モデルの能力獲得過程. 第21回YANSシンポジウム (YANS2026), August 2026.
吉田 遥音, 工藤 慧音, 松田 陵佑, 清水 伸幸, 鈴木 潤. 長尺動画生成タスクにおけるプロンプトと動画間のシーン対応を考慮した自動評価. 第21回YANSシンポジウム (YANS2026), August 2026.
羽根田 賢和, 大竹 孝樹, 塙 一晃, 樋口 泰地, 山口 学. 社会的関心の尺度として Wikipedia 閲覧数を利用した Web メディアの記事人気予測. 2026年度人工知能学会全国大会 (第40回), June 2026.
渡邉 天照, 羽根田 賢和, 坂口 慶祐. TRPGゲームマスターAIの迎合性がプレイヤー体験に及ぼす影響. 2026年度人工知能学会全国大会 (第40回), June 2026.
松﨑 孝介, 谷口 雅弥, 坂口 慶祐. 日本語方言の音声認識における学習データの規模と特性. 言語処理学会第32回年次大会 (NLP2026), March 2026.
Hong Hai Ngo, Yunmeng Li, Steven Coyne, Kamila Szelag-Levine, 坂口 慶祐. Wav2Karaoke: A Pipeline to Align Notes and Lyrics from a Song. 言語処理学会第32回年次大会 (NLP2026), March 2026.
野末 慎之介, 松林 優一郎, 藤井 諒, 岸波 洋介, 森下 睦, 坂口 慶祐. 項の復元は必要か？日英機械翻訳における省略された項の扱いの分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
佐々木 睦史, 中山 功太, 宮尾 祐介, 大関 洋平, 磯沼 大. 排他的逆学習. 言語処理学会第32回年次大会 (NLP2026), March 2026.
澤田 紘希, 佐々木 睦史, 坂口 慶祐, Benjamin Heinzerling. 言語モデルにおける共参照解決を担う注意ヘッドの分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
神林 励旺, 稲葉 達郎, Benjamin Heinzerling, 坂口 慶祐, 乾 健太郎, 磯沼 大. 画像生成学習が視覚理解能力へ与える影響の分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
原知正, 栗田宙人, 今泉允聡, 乾健太郎, 横井祥. なぜ平均プーリングはうまく動くのか？テキスト埋め込みの二次統計量の崩壊の定量化. 言語処理学会第32回年次大会 (NLP2026), March 2026.
木谷頼斗, 大橋諭貴, 佐藤宏亮, 鴨田豪, 高橋良允, 山本悠士, 塩野大輝, 坂口慶祐, 小林悟郎. Attention Sink および Massive Activation の発生機序の分解. 言語処理学会第32回年次大会 (NLP2026), March 2026.
大橋 諭貴, 木谷 頼斗, 佐藤 宏亮, 高橋 良允, 鴨田 豪, 山本 悠士, 塩野 大輝, 坂口 慶祐, 小林 悟郎. 注意機構における Attention Sink のバイアス項的解釈. 言語処理学会第32回年次大会 (NLP2026), March 2026
猪股和樹, 吉田遥音, 塩野大輝, 赤間怜奈, 坂口慶祐. 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. 言語処理学会第32回年次大会 (NLP2026), March 2026.
高橋良允, 稲葉達郎, 乾健太郎, Benjamin Heinzerling. TopK Language Models. 言語処理学会第32回年次大会 (NLP2026), March 2026.
坂田将樹, Benjamin Heinzerling, 伊藤拓海, 横井祥, 乾健太郎. 言語モデルに線形に内在する階層的概念. 言語処理学会第32回年次大会 (NLP2026), March 2026.
青木洋一, 村上聡一朗, 本多右京, 加藤明彦. 広告文生成における多様性と広告品質の関係. 言語処理学会第32回年次大会 (NLP2026), March 2026.
Other Publications and Presentations
【その他の出版物および発表】
吉村 賢人, 高見 享佑, 古橋 萌々香, 中山 功太 , 児玉 貴志. 小学校漢字学習におけるLLM生成物語の「違和感」の類型化とLLM間の比較. 第21回YANSシンポジウム (YANS2026), August 2026.
松﨑 孝介, 谷口 雅弥. 「待つ」は一瞬か数時間か：LLMによるイベント持続時間の評価. 第21回YANSシンポジウム (YANS2026), August 2026.
工藤 慧音, 山口健史, 坂口 慶祐. nanochat-jp: 小規模日本語特化言語モデル統合研究基盤. 第21回YANSシンポジウム (YANS2026), August 2026.
佐々木 睦史, 葉 夢宇, 趙 羽風, 工藤 慧音, 坂口 慶祐, 鈴木 潤. 一様拡散言語モデルの能力獲得過程. 第21回YANSシンポジウム (YANS2026), August 2026.
吉田 遥音, 工藤 慧音, 松田 陵佑, 清水 伸幸, 鈴木 潤. 長尺動画生成タスクにおけるプロンプトと動画間のシーン対応を考慮した自動評価. 第21回YANSシンポジウム (YANS2026), August 2026.
羽根田 賢和, 大竹 孝樹, 塙 一晃, 樋口 泰地, 山口 学. 社会的関心の尺度として Wikipedia 閲覧数を利用した Web メディアの記事人気予測. 2026年度人工知能学会全国大会 (第40回), June 2026.
渡邉 天照, 羽根田 賢和, 坂口 慶祐. TRPGゲームマスターAIの迎合性がプレイヤー体験に及ぼす影響. 2026年度人工知能学会全国大会 (第40回), June 2026.
松﨑 孝介, 谷口 雅弥, 坂口 慶祐. 日本語方言の音声認識における学習データの規模と特性. 言語処理学会第32回年次大会 (NLP2026), March 2026.
Hong Hai Ngo, Yunmeng Li, Steven Coyne, Kamila Szelag-Levine, 坂口 慶祐. Wav2Karaoke: A Pipeline to Align Notes and Lyrics from a Song. 言語処理学会第32回年次大会 (NLP2026), March 2026.
野末 慎之介, 松林 優一郎, 藤井 諒, 岸波 洋介, 森下 睦, 坂口 慶祐. 項の復元は必要か？日英機械翻訳における省略された項の扱いの分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
佐々木 睦史, 中山 功太, 宮尾 祐介, 大関 洋平, 磯沼 大. 排他的逆学習. 言語処理学会第32回年次大会 (NLP2026), March 2026.
澤田 紘希, 佐々木 睦史, 坂口 慶祐, Benjamin Heinzerling. 言語モデルにおける共参照解決を担う注意ヘッドの分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
神林 励旺, 稲葉 達郎, Benjamin Heinzerling, 坂口 慶祐, 乾 健太郎, 磯沼 大. 画像生成学習が視覚理解能力へ与える影響の分析. 言語処理学会第32回年次大会 (NLP2026), March 2026.
原知正, 栗田宙人, 今泉允聡, 乾健太郎, 横井祥. なぜ平均プーリングはうまく動くのか？テキスト埋め込みの二次統計量の崩壊の定量化. 言語処理学会第32回年次大会 (NLP2026), March 2026.
木谷頼斗, 大橋諭貴, 佐藤宏亮, 鴨田豪, 高橋良允, 山本悠士, 塩野大輝, 坂口慶祐, 小林悟郎. Attention Sink および Massive Activation の発生機序の分解. 言語処理学会第32回年次大会 (NLP2026), March 2026.
大橋 諭貴, 木谷 頼斗, 佐藤 宏亮, 高橋 良允, 鴨田 豪, 山本 悠士, 塩野 大輝, 坂口 慶祐, 小林 悟郎. 注意機構における Attention Sink のバイアス項的解釈. 言語処理学会第32回年次大会 (NLP2026), March 2026
猪股和樹, 吉田遥音, 塩野大輝, 赤間怜奈, 坂口慶祐. 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. 言語処理学会第32回年次大会 (NLP2026), March 2026.
高橋良允, 稲葉達郎, 乾健太郎, Benjamin Heinzerling. TopK Language Models. 言語処理学会第32回年次大会 (NLP2026), March 2026.
坂田将樹, Benjamin Heinzerling, 伊藤拓海, 横井祥, 乾健太郎. 言語モデルに線形に内在する階層的概念. 言語処理学会第32回年次大会 (NLP2026), March 2026.
青木洋一, 村上聡一朗, 本多右京, 加藤明彦. 広告文生成における多様性と広告品質の関係. 言語処理学会第32回年次大会 (NLP2026), March 2026.
Thesis
【学位論文】
[Master's Thesis] 吉田遥音: Internal Representations of Diagrams in Vision Language Models 2026.03
[Master's Thesis] 羽根田賢和: 短歌の「味」を測る: 言語モデルを用いた短歌の定量的評価 2026.03
[Master's Thesis] 松﨑 孝介: 低資源言語の音声認識における学習データ規模と性能の定量的評価 2026.03
[Master's Thesis] 原 知正: An Analysis of Spatial Spread Collapse in Word Embedding Sets for Text Embeddings. 2026.03
[Master's Thesis] 中野 雄斗: 説得対話エージェントの拡張戦略設計と意図変化を用いた自動評価手法の検証 2026.03
Thesis
【学位論文】
[Master's Thesis] 吉田遥音: Internal Representations of Diagrams in Vision Language Models 2026.03
[Master's Thesis] 羽根田賢和: 短歌の「味」を測る: 言語モデルを用いた短歌の定量的評価 2026.03
[Master's Thesis] 松﨑 孝介: 低資源言語の音声認識における学習データ規模と性能の定量的評価 2026.03
[Master's Thesis] 原 知正: An Analysis of Spatial Spread Collapse in Word Embedding Sets for Text Embeddings. 2026.03
[Master's Thesis] 中野 雄斗: 説得対話エージェントの拡張戦略設計と意図変化を用いた自動評価手法の検証 2026.03
Thesis
【学位論文】
[Master's Thesis] 吉田遥音: Internal Representations of Diagrams in Vision Language Models 2026.03
[Master's Thesis] 羽根田賢和: 短歌の「味」を測る: 言語モデルを用いた短歌の定量的評価 2026.03
[Master's Thesis] 松﨑 孝介: 低資源言語の音声認識における学習データ規模と性能の定量的評価 2026.03
[Master's Thesis] 原 知正: An Analysis of Spatial Spread Collapse in Word Embedding Sets for Text Embeddings. 2026.03
[Master's Thesis] 中野 雄斗: 説得対話エージェントの拡張戦略設計と意図変化を用いた自動評価手法の検証 2026.03
