2026OtherNLP 2026
MoEアーキテクチャによる破滅的忘却の抑制効果の評価
Hinata Sugimoto, Jaesung Lee, Ko Yoshida, Jun Suzuki
LLMMoECatastrophic Forgettingpretraining
We evaluate the effectiveness of MoE architecture in suppressing catastrophic forgetting during pretraining.
Publications
Publication list sorted by year with filters for type and tags.
2025PreprintarXiv
An Open and Reproducible Deep Research Agent for Long-Form Question Answering
Ikuya Yamada, Wataru Ikeda, Ko Yoshida, Mengyu Ye, Hinata Sugimoto, Masatoshi Suzuki, Hisanori Ozaki, Jun Suzuki
RAGDeep Research Agent
We present an open deep research system for long-form question answering, selected as a winning system in the text-to-text track of the MMU-RAG competition at NeurIPS 2025.