OpenReview
.net
OpenReview
.net
Login
OpenReview
.net
Login
Hakaze Cho
Researcher, Tohoku University
Researcher, Center for Advanced Intelligence Project, RIKEN
PhD student, Japan Advanced Institute of Science and Technology
Joined
January 2024
Names
Hakaze Cho
(Preferred)
,
Yufeng Zhao
,
趙羽風
Emails
****@jaist.ac.jp
(Confirmed)
,
****@outlook.com
(Confirmed)
,
****@riken.jp
(Confirmed)
Personal Links
Homepage
Google Scholar
DBLP
ORCID
Semantic Scholar
ACL Anthology
Career & Education History
Researcher
Tohoku University
(tohoku.ac.jp)
2026
–
2029
Researcher
Center for Advanced Intelligence Project,
RIKEN
(riken.jp)
2026
–
2029
PhD student
Japan Advanced Institute of Science and Technology
(jaist.ac.jp)
2023
–
2026
MS student
Beijing Institute of Technology
(bit.edu.cn)
2021
–
2023
Undergrad student
Beijing Institute of Technology
(bit.edu.cn)
2017
–
2021
Advisors, Relations & Conflicts
Coauthor
Haolin Yang
Present
Coauthor
Gouki Minegishi
Present
Advisee
Rui Li
2025
–
Present
Postdoc Advisor
Kentaro Inui
2026
–
2029
PhD Advisor
Naoya Inoue
2023
–
2026
Internship Advisor
Brian M. Kurkoski
2024
–
2025
Expertise
In-context Learning
2023
–
Present
Interpretability
2023
–
Present
Representation Learning
2021
–
Present
Sentiment Analysis
2021
–
2022
Publications
Task Vectors, Learned Not Extracted: Performance Gains and Mechanistic Insights
Haolin Yang
,
Hakaze Cho
,
Kaize Ding
,
Naoya Inoue
ICLR 2026 Poster
Readers:
Everyone
Localizing Task Recognition and Task Learning in In-Context Learning via Attention Head Analysis
Haolin Yang
,
Hakaze Cho
,
Naoya Inoue
ICLR 2026 Poster
Readers:
Everyone
Mechanism of Task-oriented Information Removal in In-context Learning
Hakaze Cho
,
Haolin Yang
,
Gouki Minegishi
,
Naoya Inoue
ICLR 2026 Poster
Readers:
Everyone
Binary Autoencoder for Mechanistic Interpretability of Large Language Models
Hakaze Cho
,
Haolin Yang
,
Brian M. Kurkoski
,
Naoya Inoue
Submitted to ICLR 2026
Readers:
Everyone
Unifying Attention Heads and Task Vectors via Hidden State Geometry in In-Context Learning
Haolin Yang
,
Hakaze Cho
,
Yiqiao Zhong
,
Naoya Inoue
NeurIPS 2025 poster
Readers:
Everyone
Unifying Attention Heads and Task Vectors via Hidden State Geometry in In-Context Learning
Haolin Yang
,
Hakaze Cho
,
Yiqiao Zhong
,
Naoya Inoue
CoRR 2025
Readers:
Everyone
Mechanistic Fine-tuning for In-context Learning
Hakaze Cho
,
Peng Luo
,
Mariko Kato
,
Rin Kaenbyou
,
Naoya Inoue
CoRR 2025
Readers:
Everyone
Affinity and Diversity: A Unified Metric for Demonstration Selection via Internal Representations
Mariko Kato
,
Hakaze Cho
,
Yoshihiro Sakai
,
Naoya Inoue
CoRR 2025
Readers:
Everyone
Understanding Token Probability Encoding in Output Embeddings
Hakaze Cho
,
Yoshihiro Sakai
,
Kenshiro Tanaka
,
Mariko Kato
,
Naoya Inoue
COLING 2025
Readers:
Everyone
Revisiting In-context Learning Inference Circuit in Large Language Models
Hakaze Cho
,
Mariko Kato
,
Yoshihiro Sakai
,
Naoya Inoue
ICLR 2025
Readers:
Everyone
View all 14 publications
Co-Authors
Akira Ishii
Brian M. Kurkoski
Gouki Minegishi
Haolin Yang
Kaize Ding
Kenshiro Tanaka
Mariko Kato
Naoya Inoue
Peng Luo
Rin Kaenbyou
Yiqiao Zhong
Yoshihiro Sakai