Eighteen papers involving researchers from the RIKEN Center for Advanced Intelligence Project (AIP) have been accepted to EMNLP 2026 (Conference on Empirical Methods in Natural Language Processing), one of the leading international conferences in natural language processing. Nine papers were accepted to the Main Conference, seven papers to Findings of EMNLP 2026, and two papers to other tracks and co-located conferences. For more information, please visit EMNLP 2026.
Main Conference:
A Robust Evaluation of Probe Robustness: Lessons for Reliable OOD Uncertainty Quantification
Joe Stacey (University of Sheffield)
Hadas Orgad (Kempner Institute at Harvard University)
Kentaro Inui (MBZUAI / Tohoku University / RIKEN)
Benjamin Heinzerling (RIKEN / Tohoku University)
Nafise Sadat Moosavi (University of Sheffield)
It’s Not RoPE That Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention
Raito Kiya (Tohoku University)
Satoki Ohashi (Tohoku University)
Kosuke Sato (Tohoku University)
Go Kamoda (SOKENDAI / NINJAL)
Ryosuke Takahashi (Tohoku University / MBZUAI / RIKEN)
Yuji Yamamoto (SOKENDAI / NINJAL)
Daiki Shiono (Tohoku University)
Keisuke Sakaguchi (Tohoku University / RIKEN)
Goro Kobayashi (Preferred Networks, Inc. / Tohoku University)
Weakening Neurons: An Input-Output Functionality in Transformers with Outsize Influence
Sebastian Gerstner (LMU Munich)
Hilal AlQuabeh (MBZUAI)
Kentaro Inui (MBZUAI / Tohoku University / RIKEN)
Hinrich Schuetze (LMU Munich)
When Queries Crowd: Diagnosing and Improving Spread-Out Regularization for Text Embeddings
Zeqi Zhou (The University of Tokyo)
Qingcheng Zeng (Northwestern University)
Rui Yang (National University of Singapore)
Jiarui Liu (Carnegie Mellon University)
Heli Qi (RIKEN AIP)
Puxuan Yu (Snowflake Inc.)
Weihao Xuan (The University of Tokyo / RIKEN AIP)
Naoto Yokoya (The University of Tokyo / RIKEN AIP)
Length and Meaning for Language Model Maps
Momose Oyama (Kyoto University / RIKEN AIP)
Ryo Kishino (Kyoto University)
Hiroaki Yamagiwa (SB Intuitions)
Hidetoshi Shimodaira (Kyoto University / RIKEN AIP)
Language Model Maps for Prompt-Response Distributions via Log-Likelihood Vectors
Momose Oyama (Kyoto University / RIKEN AIP)
Yusuke Takase (Kyoto University)
Hidetoshi Shimodaira (Kyoto University / RIKEN AIP)
Reasoning Depth and Environment Complexity: A Controlled Study of RLVR Data Allocation across Logical Reasoning Tasks
Yihua Zhu (Kyoto University / NII LLMC)
Qianying Liu (NII LLMC)
Fei Cheng (Kyoto University)
Jiaxin Wang (Kyoto University)
Akiko Aizawa (The University of Tokyo / NII LLMC)
Sadao Kurohashi (NII LLMC)
Hidetoshi Shimodaira (Kyoto University / RIKEN AIP)
Cross-Lingual Representation Alignment by Token-Level Optimal Transport in a Language-Agnostic Space
Taisei Yamamoto (The University of Tokyo / RIKEN AIP)
Ryoma Kumon (The University of Tokyo / RIKEN AIP)
Danushka Bollegala (University of Liverpool)
Hitomi Yanaka (The University of Tokyo / RIKEN AIP / Tohoku University)
(How) Do MLLMs Report Bistable Images Like Humans?
Ryota Takatsuki (University of Sussex)
Tomoki Doi (The University of Tokyo / RIKEN AIP)
Amane Watahiki (The University of Tokyo / RIKEN AIP)
Anil K. Seth (University of Sussex)
Hitomi Yanaka (The University of Tokyo / RIKEN AIP / Tohoku University)
Findings:
Can Dialects Be Steered Like Languages? Sparse Neurons and Distributed Directions in Arabic LLMs
Kareem Elozeiri (MBZUAI)
Mervat Abassy (MBZUAI)
Omar Kallas (MBZUAI)
Fahim Dalvi (QCRI, HBKU)
Preslav Nakov (MBZUAI)
Kentaro Inui (MBZUAI / Tohoku University / RIKEN)
Nadir Durrani (QCRI, HBKU)
LITERARYBIGFIVE: Author-Personalized Text Generation in a Unified Interpretable Space
Jinghui Zhang (MBZUAI)
Lang Gao (MBZUAI)
Ao Li (Shandong University)
Mingzhe Li (Independent Researcher)
Ruihong Zeng (MBZUAI)
Zirui Song (MBZUAI)
Kentaro Inui (MBZUAI / Tohoku University / RIKEN)
Xiuying Chen (MBZUAI)
Sparse Long Memory and Transition Bottlenecks Shape Positional Recall in Mamba Language Models
Muhammad Cendekia Airlangga (MBZUAI)
Hilal AlQuabeh (MBZUAI)
Munachiso Samuel Nwadike (MBZUAI)
Kentaro Inui (MBZUAI / Tohoku University / RIKEN)
Domain Mixture Design via Log-Likelihood Differences for Aligning Language Models with a Target Model
Ryo Kishino (Kyoto University)
Riku Shiomi (Kyoto University)
Hiroaki Yamagiwa (SB Intuitions)
Momose Oyama (Kyoto University / RIKEN AIP)
Hidetoshi Shimodaira (Kyoto University / RIKEN AIP)
Harmless Yet Harmful: Neutral Prompting Attacks for Stealthy Hallucination Steering in Agent Skills
Chia-Yi Hsu (National Yang Ming Chiao Tung University)
Chia-Mu Yu (National Yang Ming Chiao Tung University)
Chun-Ying Huang (National Yang Ming Chiao Tung University)
Jun Sakuma (RIKEN AIP / Institute of Science Tokyo)
A Single Suffix to Break Them All: Basin-Aware Jailbreaks for Merged Model Families
Yu Zhe (RIKEN AIP)
Yixin Tan (Institute of Science Tokyo)
Junhao Wei (RIKEN AIP / Institute of Science Tokyo)
Chen Wang (Zhejiang University)
Shared Low-rank Basis Factorization for Data-free Mixture-of-Experts Compression
Tianxiao Cao (Kyoto University)
Jiahe Shao (The University of Tokyo)
Yuning Qiu (RIKEN AIP)
Kyohei Atarashi (Kyoto University)
Hisashi Kashima (Kyoto University)
Qibin Zhao (RIKEN AIP)
System Demonstration Track:
CAT: A CounterArgument Tutoring System based on Logical Structure Analysis
Wenzhi Wang (Tohoku University / RIKEN)
Thiemo Wambsganss (Bern University of Applied Sciences)
Chenyang Wang (Bern University of Applied Sciences)
Paul Reisert (Beyond Reason)
Léane Wettstein (Bern University of Applied Sciences)
Katja Pott (Bern University of Applied Sciences)
Kenshi Yamaguchi (RIKEN)
Machi Shimmei (Tohoku University / RIKEN)
Roman Rietsche (Bern University of Applied Sciences)
Surawat Pothong (JAIST)
Naoya Inoue (JAIST)
Kentaro Inui (MBZUAI / Tohoku University / RIKEN)
WMT 2026 (Conference on Machine Translation)
Doc2FRC: Length-Consistent Document-Level Machine Translation via Fixed-Range Chunking
Xiaotian Wang (The University of Tokyo / RIKEN AIP)
Youyuan Lin (Kyoto University)
Zhan Shen (The University of Tokyo)
Hitomi Yanaka (The University of Tokyo / RIKEN AIP / Tohoku University)
