Shun Kiyono

39 papers A* 2B 1C 1Misc 1Journal 20Unranked 14
YearRankTypeTitle / Venue / Authors
2026 J jnl
Pre-training LLM without Learning Rate Decay Enhances Supervised Fine-Tuning.
CoRR
Kazuki Yano, Shun Kiyono, Sosuke Kobayashi, Sho Takase, Jun Suzuki
2025 J jnl
Efficient Construction of Model Family through Progressive Training Using Model Expansion.
CoRR
Kazuki Yano, Sho Takase, Sosuke Kobayashi, Shun Kiyono, Jun Suzuki
2025 — conf
Large Vocabulary Size Improves Large Language Models.
ACL (Findings)
Sho Takase, Ryokan Ri, Shun Kiyono, Takuya Kato
2024 J jnl
Large Vocabulary Size Improves Large Language Models.
CoRR
Sho Takase, Ryokan Ri, Shun Kiyono, Takuya Kato
2024 J jnl
Self-Translate-Train: A Simple but Strong Baseline for Cross-lingual Transfer of Large Language Models.
CoRR
Ryokan Ri, Shun Kiyono, Sho Takase
2023 — conf
B2T Connection: Serving Stability and Performance in Deep Transformers.
ACL (Findings)
Sho Takase, Shun Kiyono, Sosuke Kobayashi, Jun Suzuki
2023 Misc conf
Bridging the Gap between Subword and Character Segmentation in Pretrained Language Models.
RANLP
Shun Kiyono, Sho Takase, Shengzhe Li, Toshinori Sato
2023 — conf
Lessons on Parameter Sharing across Layers in Transformers.
SustaiNLP
Sho Takase, Shun Kiyono
2023 J jnl
Spike No More: Stabilizing the Pre-training of Large Language Models.
CoRR
Sho Takase, Shun Kiyono, Sosuke Kobayashi, Jun Suzuki
2022 J jnl
Diverse Lottery Tickets Boost Ensemble from a Single Pretrained Model.
CoRR
Sosuke Kobayashi, Shun Kiyono, Jun Suzuki, Kentaro Inui
2022 — conf
NT5 at WMT 2022 General Translation Task.
WMT
Makoto Morishita, Keito Kudo, Yui Oka, Katsuki Chousa, Shun Kiyono, Sho Takase, Jun Suzuki
2022 J jnl
On Layer Normalizations and Residual Connections in Transformers.
CoRR
Sho Takase, Shun Kiyono, Sosuke Kobayashi, Jun Suzuki
2021 J jnl
Lessons on Parameter Sharing across Layers in Transformers.
CoRR
Sho Takase, Shun Kiyono
2021 — conf
Pseudo Zero Pronoun Resolution Improves Zero Anaphora Resolution.
EMNLP (1)
Ryuto Konno, Shun Kiyono, Yuichiroh Matsubayashi, Hiroki Ouchi, Kentaro Inui
2021 J jnl
Pseudo Zero Pronoun Resolution Improves Zero Anaphora Resolution.
CoRR
Ryuto Konno, Shun Kiyono, Yuichiroh Matsubayashi, Hiroki Ouchi, Kentaro Inui
2021 — conf
Rethinking Perturbations in Encoder-Decoders for Fast Training.
NAACL-HLT
Sho Takase, Shun Kiyono
2021 J jnl
Rethinking Perturbations in Encoder-Decoders for Fast Training.
CoRR
Sho Takase, Shun Kiyono
2021 — conf
SHAPE : Shifted Absolute Position Embedding for Transformers.
EMNLP (1)
Shun Kiyono, Sosuke Kobayashi, Jun Suzuki, Kentaro Inui
2021 J jnl
SHAPE: Shifted Absolute Position Embedding for Transformers.
CoRR
Shun Kiyono, Sosuke Kobayashi, Jun Suzuki, Kentaro Inui
2020 — conf
A Self-Refinement Strategy for Noise Reduction in Grammatical Error Correction.
EMNLP (Findings)
Masato Mita, Shun Kiyono, Masahiro Kaneko, Jun Suzuki, Kentaro Inui
2020 J jnl
A Self-Refinement Strategy for Noise Reduction in Grammatical Error Correction.
CoRR
Masato Mita, Shun Kiyono, Masahiro Kaneko, Jun Suzuki, Kentaro Inui
2020 B conf
An Empirical Study of Contextual Data Augmentation for Japanese Zero Anaphora Resolution.
COLING
Ryuto Konno, Yuichiroh Matsubayashi, Shun Kiyono, Hiroki Ouchi, Ryo Takahashi, Kentaro Inui
2020 J jnl
An Empirical Study of Contextual Data Augmentation for Japanese Zero Anaphora Resolution.
CoRR
Ryuto Konno, Yuichiroh Matsubayashi, Shun Kiyono, Hiroki Ouchi, Ryo Takahashi, Kentaro Inui
2020 — conf
ESPnet-ST: All-in-One Speech Translation Toolkit.
ACL (demo)
Hirofumi Inaguma, Shun Kiyono, Kevin Duh, Shigeki Karita, Nelson Yalta, Tomoki Hayashi, Shinji Watanabe
2020 J jnl
ESPnet-ST: All-in-One Speech Translation Toolkit.
CoRR
Hirofumi Inaguma, Shun Kiyono, Kevin Duh, Shigeki Karita, Nelson Enrique Yalta Soplin, Tomoki Hayashi, Shinji Watanabe
2020 A* conf
Encoder-Decoder Models Can Benefit from Pre-trained Masked Language Models in Grammatical Error Correction.
ACL
Masahiro Kaneko, Masato Mita, Shun Kiyono, Jun Suzuki, Kentaro Inui
2020 J jnl
Encoder-Decoder Models Can Benefit from Pre-trained Masked Language Models in Grammatical Error Correction.
CoRR
Masahiro Kaneko, Masato Mita, Shun Kiyono, Jun Suzuki, Kentaro Inui
2020 J jnl
Massive Exploration of Pseudo Data for Grammatical Error Correction.
IEEE ACM Trans. Audio Speech Lang. Process.
Shun Kiyono, Jun Suzuki, Tomoya Mizumoto, Kentaro Inui
2020 — conf
Tohoku-AIP-NTT at WMT 2020 News Translation Task.
WMT@EMNLP
Shun Kiyono, Takumi Ito, Ryuto Konno, Makoto Morishita, Jun Suzuki
2019 — conf
An Empirical Study of Incorporating Pseudo Data into Grammatical Error Correction.
EMNLP/IJCNLP (1)
Shun Kiyono, Jun Suzuki, Masato Mita, Tomoya Mizumoto, Kentaro Inui
2019 J jnl
An Empirical Study of Incorporating Pseudo Data into Grammatical Error Correction.
CoRR
Shun Kiyono, Jun Suzuki, Masato Mita, Tomoya Mizumoto, Kentaro Inui
2019 — conf
ESPnet How2 Speech Translation System for IWSLT 2019: Pre-training, Knowledge Distillation, and Going Deeper.
IWSLT
Hirofumi Inaguma, Shun Kiyono, Nelson Enrique Yalta Soplin, Jun Suzuki, Kevin Duh, Shinji Watanabe
2019 — conf
Effective Adversarial Regularization for Neural Machine Translation.
ACL (1)
Motoki Sato, Jun Suzuki, Shun Kiyono
2019 A* conf
Mixture of Expert/Imitator Networks: Scalable Semi-Supervised Learning Framework.
AAAI
Shun Kiyono, Jun Suzuki, Kentaro Inui
2019 J jnl
Riposte! A Large Corpus of Counter-Arguments.
CoRR
Paul Reisert, Benjamin Heinzerling, Naoya Inoue, Shun Kiyono, Kentaro Inui
2018 J jnl
Mixture of Expert/Imitator Networks: Scalable Semi-supervised Learning Framework.
CoRR
Shun Kiyono, Jun Suzuki, Kentaro Inui
2018 C conf
Reducing Odd Generation from Neural Headline Generation.
PACLIC
Shun Kiyono, Sho Takase, Jun Suzuki, Naoaki Okazaki, Kentaro Inui, Masaaki Nagata
2018 — conf
Unsupervised Token-wise Alignment to Improve Interpretation of Encoder-Decoder Models.
BlackboxNLP@EMNLP
Shun Kiyono, Sho Takase, Jun Suzuki, Naoaki Okazaki, Kentaro Inui, Masaaki Nagata
2017 J jnl
Source-side Prediction for Neural Headline Generation.
CoRR
Shun Kiyono, Sho Takase, Jun Suzuki, Naoaki Okazaki, Kentaro Inui, Masaaki Nagata