BL.

NLP researcher · Yokohama, Japan

Boxuan Lyu呂 博軒

PhD student · Institute of Science Tokyo
Okumura–Funakoshi Lab

Making language generation more reliable and efficient.

I study machine translation, with a focus on decoding, evaluation, and learning from model-generated feedback.

Latest · 2026

Iterative MBR distillation · EMNLP Findings
NLP 2026 優秀賞 · SIG-AGI 研究会優秀賞

Selected research

EMNLP 2026Findings

Is Human Annotation Necessary?

Iterative MBR Distillation for Error Span Detection in Machine Translation

Boxuan Lyu, Haiyue Song, Zhi Qu

Paper ↗Code ↗

ACL 2025Main conference

Unveiling the Power of Source

Source-based Minimum Bayes Risk Decoding for Neural Machine Translation

Boxuan Lyu, Hidetaka Kamigaito, Kotaro Funakoshi, Manabu Okumura

Paper ↗Code ↗

PREPRINT2026 · Under review

Quit While You're Ahead

Quit for Efficient Candidate Generation in Machine Translation Reranking

Guangyu Chen, Boxuan Lyu, Hidetaka Kamigaito, Kotaro Funakoshi, Manabu Okumura

Paper ↗

PREPRINT2026 · Under review

Routing by Analogy

kNN-Augmented Expert Assignment for Mixture-of-Experts

Boxuan Lyu, Soichiro Murakami, Hidetaka Kamigaito, Peinan Zhang

Paper ↗

PREPRINT2025 · Under review

Minimum Bayes Risk Decoding for Error Span Detection in Reference-Free Automatic Machine Translation Evaluation

Boxuan Lyu, Haiyue Song, Hidetaka Kamigaito, Chenchen Ding, Hideki Tanaka, Masao Utiyama, Kotaro Funakoshi, Manabu Okumura

Paper ↗Code ↗

PREPRINT2026 · Under review

Rewrite to Translate, Translate to Reward

Reinforcement Learning for Source Rewriting in Machine Translation

Boxuan Lyu, Haiyue Song, Zhi Qu, Hidetaka Kamigaito, Kotaro Funakoshi, Manabu Okumura

Paper ↗

SIG-AGI 2026第32回汎用人工知能研究会

言語と行動を同時生成する2元時系列生成モデル

川上 晴也, 呂 博軒, 船越 孝太郎, 奥村 学 · 研究会優秀賞

Paper ↗

NLP 2026言語処理学会第32回年次大会

類推に基づくエキスパート割当

kNN検索を用いたMixture-of-Experts

呂 博軒, 村上 聡一郎, 上垣外 英剛, 張 培楠 · 優秀賞

Paper ↗

YANS 2025第20回言語処理若手シンポジウム

生成型自動評価指標のための最小ベイズリスク復号

呂 博軒, 宋 海越, 上垣外 英剛, 田中 英輝, 内山 将夫, 船越 孝太郎, 奥村 学

NLP 2025言語処理学会第31回年次大会

漫画話者認識における VLM の有効性

呂 博軒, 能地 宏

Paper ↗

YANS 2024第19回言語処理若手シンポジウム

なぜ機械翻訳モデルの推定確率と人手評価の相関係数が低いのか？

原因分析と改善策の検討

呂 博軒, 上垣外 英剛, 船越 孝太郎, 奥村 学

Experience

CyberAgent AI LabAug 2025 — Present · Research internship

NICT UCRI ASTRECJun 2025 — Mar 2026 · Research internship

Earlier experience

Mantra Inc.Oct 2024 — Jun 2025 · Research internship

NEC R&DAug 2024 — Oct 2024 · Research internship

ELYZA, Inc.May 2023 — May 2024 · AI engineering internship

Education

Institute of Science TokyoApr 2025 — Present · PhD, EngineeringSupervisor · Kotaro Funakoshi

Institute of Science TokyoApr 2023 — Mar 2025 · Master, EngineeringSupervisor · Kotaro Funakoshi

Jiangnan UniversitySep 2018 — Jun 2022 · Bachelor, ScienceInformation and Computing Science

Visitors around the world

An aggregate, privacy-aware view of where this site is read. No personal profiles or precise locations are stored.

Visits —
Countries / regions —

This month

Loading map…

Preparing the first visitor report.