Goro Kobayashi, Tatsuki Kuribayashi, Sho Yokoi, Kentaro Inui. Transformer Language Models Handle Word Frequency in Prediction Head. CoRR. 2023. abs/2305.18294
Goro Kobayashi, Tatsuki Kuribayashi, Sho Yokoi, Kentaro Inui. Feed-Forward Blocks Control Contextualization in Masked Language Models. CoRR. 2023. abs/2302.00456
Goro Kobayashi, Tatsuki Kuribayashi, Sho Yokoi, Kentaro Inui. Transformer Language Models Handle Word Frequency in Prediction Head. ACL (Findings). 2023. 4523-4535
Momose Oyama, Sho Yokoi, Hidetoshi Shimodaira. Norm of word embedding encodes information gain. 2022
Hiroaki Yamagiwa, Sho Yokoi, Hidetoshi Shimodaira. Improving word mover's distance by leveraging self-attention matrix. CoRR. 2022. abs/2211.06229
