Yoshihiro Sakai

11 papers A* 1B 1Journal 5Unranked 4
YearRankTypeTitle / Venue / Authors
2025 J jnl
Affinity and Diversity: A Unified Metric for Demonstration Selection via Internal Representations.
CoRR
Mariko Kato, Hakaze Cho, Yoshihiro Sakai, Naoya Inoue
2025 A* conf
Revisiting In-context Learning Inference Circuit in Large Language Models.
ICLR
Hakaze Cho, Mariko Kato, Yoshihiro Sakai, Naoya Inoue
2025 — conf
Token-based Decision Criteria Are Suboptimal in In-context Learning.
NAACL (Long Papers)
Hakaze Cho, Yoshihiro Sakai, Mariko Kato, Kenshiro Tanaka, Akira Ishii, Naoya Inoue
2025 B conf
Understanding Token Probability Encoding in Output Embeddings.
COLING
Hakaze Cho, Yoshihiro Sakai, Kenshiro Tanaka, Mariko Kato, Naoya Inoue
2024 J jnl
NoisyICL: A Little Noise in Model Parameters Calibrates In-context Learning.
CoRR
Yufeng Zhao, Yoshihiro Sakai, Naoya Inoue
2024 J jnl
Revisiting In-context Learning Inference Circuit in Large Language Models.
CoRR
Hakaze Cho, Mariko Kato, Yoshihiro Sakai, Naoya Inoue
2024 J jnl
Token-based Decision Criteria Are Suboptimal in In-context Learning.
CoRR
Hakaze Cho, Yoshihiro Sakai, Mariko Kato, Kenshiro Tanaka, Akira Ishii, Naoya Inoue
2024 J jnl
Understanding Token Probability Encoding in Output Embeddings.
CoRR
Hakaze Cho, Yoshihiro Sakai, Kenshiro Tanaka, Mariko Kato, Naoya Inoue
2018 — conf
An Improved structure of Acoustic Echo Cancellation Using Blind Source Separation.
ISPACS
Yoshihiro Sakai
2014 — conf
The acoustic echo cancelation using blind source separation to reduce double-talk interference.
IWAENC
Yoshihiro Sakai, Muhammad Tahir Akhtar
2013 — conf
The performance of the acoustic echo cancelation using blind source separation to reduce double-talk interference.
ISPACS
Yoshihiro Sakai, Muhammad Tahir Akhtar