News

News Release

2026/07/22

20 Papers by National Institute of Informatics Researchers Accepted at ACL 2026, the Top Conference in Natural Language Processing

　Papers involving researchers of the National Institute of Informatics (NII; Director-General: Sadao Kurohashi; Chiyoda-ku, Tokyo), part of the Inter-University Research Institute Corporation, Research Organization of Information and Systems, were accepted at The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), the top conference in the field of natural language processing and computational linguistics. A total of 20 papers were accepted, two of which received the Best Theme Paper and Outstanding Paper awards.

Summary

At ACL 2026 (The 64th Annual Meeting of the Association for Computational Linguistics; July 2-7, 2026, San Diego, USA), widely regarded as the world's foremost international conference in natural language processing and computational linguistics, a total of 20 papers co-authored by NII researchers were accepted. Two of these received awards recognizing outstanding work.

Best Theme Paper：
"Mapping the Circumplex of Affect: Geometric Analysis of Emotion Representations via Hyperspherical Contrastive Learning"
Yusuke Yamauchi (University of Tokyo / Research Assistant, Aizawa Laboratory, NII); Akiko Aizawa (Professor, Digital Content and Media Sciences Research Division, NII)
[https://www.nii.ac.jp/en/news/award/2026/0707-1.html]
Outstanding Paper：
"CxMP: A Linguistic Minimal-Pair Benchmark for Evaluating Constructional Understanding in Language Models"
Miyu Oba (Nara Institute of Science and Technology / former Research Assistant, Sugawara Laboratory, NII); Saku Sugawara (Assistant Professor, Digital Content and Media Sciences Research Division, NII)
[https://www.nii.ac.jp/en/news/award/2026/0707-2.html]

In addition, the paper "The Imperfective Paradox in Large Language Models", co-authored by Yusuke Miyao (Project Professor, Research and Development Center for Large Language Models, NII / University of Tokyo), was selected for a Best Paper Award, the conference's highest honor.
ACL 2026 received a record 12,148 submissions (up approximately 45% year on year). With 2,296 papers accepted to the main conference (acceptance rate 18.9%) and 2,163 accepted to Findings (17.8%), competition was intense -- making these 20 acceptances and two awards a notable achievement.

Background and Challenges

With the rapid spread of large language models (LLMs), there is growing importance in research that explains "why a model produces a given answer" (explainability), as well as in methods for rigorously evaluating language understanding. ACL 2026 featured "Explainability of NLP Models" as a special theme, and NII's award-winning papers sit at the core of this trend.

Social Significance and Future Outlook

These achievements form a foundation for realizing trustworthy AI. With the Research and Development Center for Large Language Models (LLMC) at its core, NII will continue to advance research and development on transparent and explainable AI.

List of Accepted Papers

Main Conference: 13 papers

NII authors are shown in bold; * indicates an award-winning paper.

C2: Scalable Rubric-Augmented Reward Modeling from Binary Preferences
Akira Kawabata, Saku Sugawara
A Dual-Task Paradigm to Investigate Sentence Comprehension Strategies in Language Models
Rei Emura, Saku Sugawara
Mapping the Circumplex of Affect: Geometric Analysis of Emotion Representations via Hyperspherical Contrastive Learning [*Best Theme Paper]
Yusuke Yamauchi, Akiko Aizawa
When Bigger Isn't Better: A Comprehensive Fairness Evaluation of Political Bias in Multi-News Summarisation
Nannan Huang, Iffat Maab, Junichi Yamagishi
MED-COREASONER: Reducing Language Disparities in Medical Reasoning via Language-Informed Co-Reasoning
Fan Gao, Sherry T. Tong, Jiwoong Sohn, Jiahao Huang, Junfeng Jiang, Ding Xia, Piyalitt Ittichaiwong, Kanyakorn Veerakanjana, Hyunjae Kim, Qingyu Chen, Edison Marrese-Taylor, Kazuma Kobayashi, Akiko Aizawa, Irene Li
Language Models Learn Universal Representations of Numbers and Here's Why You Should Care
Michal Štefánik, Timothee Mickus, Marek Kadlčík, Bertram Højer, Michal Spiegel, Raúl Vázquez, Aman Sinha, Josef Kuchař, Philipp Mondorf, Pontus Stenetorp
Attribution, Citation, and Quotation: A Survey of Evidence-based Text Generation with Large Language Models
Tobias Schreieder, Tim Schopf, Michael Färber
Label Effects: Shared Heuristic Reliance in Trust Assessment by Humans and LLM-as-a-Judge
Xin Sun, Di Wu, Sijing Qin, Isao Echizen, Abdallah El Ali, Saku Sugawara
Memorization, Emergence, and Explaining Reversal Failures: A Controlled Study of Relational Semantics in LLMs
Yihua Zhu, Qianying Liu, Jiaxin Wang, Fei Cheng, Chaoran Liu, Akiko Aizawa, Sadao Kurohashi, Hidetoshi Shimodaira
An Existence Proof for Neural Language Models That Can Explain Garden-Path Effects via Surprisal
Ryo Yoshida, Shinnosuke Isono, Taiga Someya, Yohei Oseki, Tatsuki Kuribayashi
The Role of Mixed-Language Documents for Multilingual Large Language Model Pretraining
Jiandong Shao, Raphael Tang, Crystina Zhang, Karin Sevegnani, Pontus Stenetorp, Jianfei Yang, Yao Lu
Refining and Reusing Annotation Guidelines for LLM Annotation
Kon Woo Kim, Jin-Dong Kim, Akiko Aizawa
CxMP: A Linguistic Minimal-Pair Benchmark for Evaluating Constructional Understanding in Language Models [*Outstanding Paper]
Miyu Oba, Saku Sugawara

Findings: 4 papers

Towards Reliable Paper Contributions Annotation in the ACL Rolling Review
Julien Aubert-Béduchaud, Florian Boudin, Akiko Aizawa, Beatrice Daille, Richard Dufour
ExaGPT: Example-Based Machine-Generated Text Detection for Human Interpretability
Ryuto Koike, Masahiro Kaneko, Ayana Niwa, Preslav Nakov, Naoaki Okazaki
Detecting Sensitive Personal Information in Japanese Pre-Training Corpora for Large Language Models
Rei Minamoto, Yusuke Oda, Daisuke Kawahara
J-Shuwa: A Large-Scale Web-Collected Japanese Sign Language-Japanese Parallel Corpus
Junwen Mo, MinhDuc Vo, Noriki Nishida, Shin'ichi Satoh, Hideki Nakayama

Industry Track: 1 paper

Aggregate vs. Personalized Judges in Business Idea Evaluation: Evidence from Expert Disagreement
Wataru Hirota, Tomoki Taniguchi, Tomoko Ohkuma, Kosuke Takahashi, Takahiro Omi, Kosuke Arima, Takuto Asakura, Chung-Chi Chen, Tatsuya Ishigaki

Student Research Workshop: 2 papers

Constructing a Japanese Verdict Prediction Dataset for Fact-Checking of LLM-Generated Texts
Miwa Masano, Hirokazu Kiyomaru, Atsushi Keyaki, Kaito Horio, Rei Minamoto, Ribeka Keyaki, Kouta Nakayama, Hideyuki Tachibana, Daisuke Kawahara
Task Assignment meets Annotator Modeling: Human-LLM Collaborative Annotation with Constraints
Kei Moriyama, Kouta Nakayama, Yukino Baba