革新知能統合研究センター

自然言語処理分野のトップカンファレンスであるThe 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)（7/27-8/1、Vienna（Austria）実施）において、AIPセンターから下記の通り、7本の論文が採択されました。

[ACL 2025] https://2025.aclweb.org/

[Main Oral and Panel Presentation]

• Mapping 1,000+ Language Models via the Log-Likelihood Vector
  Momose Oyama (Kyoto University / RIKEN AIP)
  Hiroaki Yamagiwa (Kyoto University)
  Yusuke Takase (Kyoto University)
  Hidetoshi Shimodaira (Kyoto University / RIKEN AIP)

[Main Poster presentation]

• Library-Like Behavior In Language Models is Enhanced by Self-Referencing Causal Cycles
  Munachiso S Nwadike (MBZUAI / RIKEN AIP)
  Zangir Iklassov (MBZUAI)
  Toluwani Aremu (MBZUAI)
  Tatsuya Hiraoka (MBZUAI / RIKEN AIP)
  Benjamin Heinzerling (RIKEN AIP / Tohoku University)
  Velibor Bojkovic (MBZUAI)
  Hilal AlQuabeh (MBZUAI / RIKEN AIP)
  Martin Takáč (MBZUAI)
  Kentaro Inui (MBZUAI / RIKEN AIP / Tohoku University)
• Quantifying Lexical Semantic Shift via Unbalanced Optimal Transport
  Ryo Kishino (Kyoto University）
  Hiroaki Yamagiwa (Kyoto University）
  Ryo Nagata (Konan University / RIKEN AIP）
  Sho Yokoi (NINJAL, Tohoku University / RIKEN AIP)
  Hidetoshi Shimodaira (Kyoto University / RIKEN AIP）
• Rubrik’s Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset
  Diana Galvan-Sosa (University of Cambridge)
  Gabrielle Gaudeau (University of Cambridge)
  Pride Kavumba (SB Intuitions)
  Yunmeng Li (Tohoku University)
  Hongyi gu (NetMind.AI)
  Zheng Yuan (The University of Sheffield)
  Keisuke Sakaguchi (Tohoku University / RIKEN AIP)
  Paula Buttery(University of Cambridge)

[Findings]

• Can Input Attributions Interpret the Inductive Reasoning Process in In-Context Learning?
  Mengyu Ye (Tohoku University)
  Tatsuki Kuribayashi (MBZUAI / Tohoku University)
  Goro Kobayashi (Tohoku University / RIKEN AIP)
  Jun Suzuki (Tohoku University / RIKEN AIP)
• On Entity Identification in Language Models
  Masaki Sakata (Tohoku University / RIKEN AIP)
  Sho Yokoi (NINJAL, Tohoku University / RIKEN AIP)
  Benjamin Heinzerling (RIKEN AIP / Tohoku University)
  Takumi Ito (Tohoku University / Langsmith Inc.)
  Kentaro Inui (MBZUAI, Tohoku University / RIKEN AIP)
• Rectifying Belief Space via Unlearning to Harness LLMs’ Reasoning
  Ayana Niwa (MBZUAI)
  Masahiro Kaneko (MBZUAI)
  Kentaro Inui (MBZUAI, Tohoku University / RIKEN AIP)

関連研究室