Yui Oka

me.png

📧 Email / 🎓 Google Scholar / 📕 OpenReview

I am a Researcher at NTT Human Informatics Laboratories, NTT Corporation, and a second-year Ph.D. student at the Language and Information Science Laboratory, Tohoku University, a member lab of the Tohoku NLP Group.

I received my Master of Engineering degree from the Nara Institute of Science and Technology (NAIST) in 2021. I was a member of the Augmented Human Communication Laboratory (AHC Lab) and was supervised by Prof. Satoshi Nakamura. My master’s research focused on neural machine translation and positional encoding. At NTT, I conduct research in Natural Language Processing, focusing on positional encoding and long-context modeling in large language models (LLMs). I also apply these theoretical insights to the development of context extension techniques for LLMs.

In recent years, I have drawn on signal processing perspectives to investigate the redesign of positional representations grounded in frequency analysis and wavelet theory.

Publications

  1. Yui Oka*, Kentaro Hanafusa*, Taku Hasegawa, Kyosuke Nishida, Kuniko Saito (* Equal contribution). “Probing Rotary Position Embeddings through Frequency Entropy.” ICLR 2026. [Openreview]
  2. Yui Oka, Itsumi Saito, Kyosuke Nishida, Kuniko Saito. “Frequency Bands in RoPE: Base Frequency and Context Length Shape the Interpolation–Extrapolation Trade-off.” ICLR 2026. [Openreview]
  3. Yui Oka, Taku Hasegawa, Kyosuke Nishida, Kuniko Saito. “Wavelet-based Positional Representation for Long Context.” ICLR 2025. [Openreview][Arxiv][Poster]
  4. Yui Oka, Tsutomu Hirao. “Implicit Sense-labeled Connective Recognition as Text Generation.” EMNLP Findings 2023. [ACL Anthology]
  5. Makoto Morishita*, Keito Kudo*, Yui Oka*, Katsuki Chousa*, Shun Kiyono, Sho Takase, Jun Suzuki (* Equal contribution). “NT5 at WMT 2022 General Translation Task.” WMT 2022. [ACL Anthology]
  6. Ryo Fukuda, Sashi Novitasari, Yui Oka, Yasumasa Kano, Yuki Yano, Yuka Ko, Hirotaka Tokuyama, Kosuke Doi, Tomoya Yanagita, Sakriani Sakti, Katsuhito Sudoh, Satoshi Nakamura. “Simultaneous Speech-to-speech Translation System with Transformer-based Incremental ASR, MT, and TTS.” O-COCOSDA 2021. [Paper] *Best Student Paper Award
  7. Ryo Fukuda, Yui Oka, Yasumasa Kano, Yuki Yano, Yuka Ko, Hirotaka Tokuyama, Kosuke Doi, Sakriani Sakti, Katsuhito Sudoh, Satoshi Nakamura. “NAIST English-to-Japanese Simultaneous Translation System for IWSLT 2021 Simultaneous Text-to-text Task.” IWSLT 2021. [ACL Anthology]
  8. Yui Oka, Katsuhito Sudoh, Satoshi Nakamura. “Using Perturbed Length-aware Positional Encoding for Non-autoregressive Neural Machine Translation.” ACL-IJCNLP SRW 2021. [Arxiv]
  9. Yui Oka, Katsuki Chousa, Katsuhito Sudoh, Satoshi Nakamura. “Incorporating Noisy Length Constraints into Transformer with Length-aware Positional Encodings.” COLING 2020. [ACL Anthology]

Biography

  • Oct 2024 – Present : Ph.D. Student, Graduate School of Information Sciences, Tohoku University
  • Jul 2024 – Present : Researcher, NTT Human Informatics Laboratories, NTT Corporation
  • Apr 2021 – Jun 2024 : Researcher, NTT Communication Science Laboratories, NTT Corporation
  • Apr 2019 – Mar 2021 : Master of Engineering, Nara Institute of Science and Technology (NAIST)

Reviewer

ICML2026, ICLR2026, ICLR2025, ACL Rolling Review (2024-present), LREC-COLING2024, EMNLP2023, COLING2022