唐突なラーメンズ

#NLP2025 の本会議と併設WSで賞を2つ受賞しました。 共著者ならびに運営・賞審査に携わられたみなさま、ありがとうございました。 [本会議] 若手奨励賞「事例ベース意思決定理論に基づく復号」 [評価指標WS 評価指標ハックshared task] best hacking award 「チームきょなら」

Hiroyuki Deguchi 出口さんのNLPコロキウムでのトークを公開しています→ 📺youtu.be/QZNZdVK7EFg 当日参加できなかったかたもぜひご覧ください！ ※ なおQA・ディスカッションは公開しておりません。 またスライドも公開いただきました。あわせてご覧ください → speakerdeck.com/de9uch1/202502…

age++; open air (Kobe)🍺

Here's a periodic reminder that you can save your readers from situations like this by using using our updated BST file, which truncates long author lists to just the first twenty github.com/acl-org/acl-st…

github.com/Unbabel/COMET/… The latest XCOMET (unbabel-comet==2.2.5) produces inconsistent results🐛 Please be careful when updating the package until the issue is resolved.

This issue has been resolved in unbabel-comet==2.2.6👍

We'll be presenting our poster on Apr 24, 3:00p.m.--5:30p.m at #ICLR2025. Please come say hi if you're interested in super-fast embedding-based pattern matching on billion-scale corpora! iclr.cc/virtual/2025/p… Code github.com/softmatcha/sof… Demo huggingface.co/spaces/softmat… #SoftMatcha

BTW, we should have Tian Tian Hainanese Chicken Rice and Song Fa Bak Kut Teh (松發肉骨茶) while we're in Singapore. 😋

🚀Announcing the 1st Workshop on Vector Databases (#VecDB@#ICML2025 in Vancouver ICML Conference)! 🎙Keynotes by Matthijs Douze, Hiroyuki Ootomo & Prashant Pandey 📢Submissions open now! (Deadline: May 27) 🔗Check the CFP for details! vecdb-ws.github.io/icml2025/index…

Meta and NVIDIA have teamed up to supercharge vector search on GPUs by integrating NVIDIA cuVS into Faiss v1.10, Meta’s open-source library for similarity search. This collaboration brings groundbreaking performance improvements: 🔹 IVF indexing: NVIDIA cuVS boosts build times

Autumn leaves...🍂

博士論文"Interpretable Neural Machine Translation from Translation to Post-Editing" がAAMT(アジア太平洋機械翻訳協会(AAMT))の長尾賞学生奨励賞に選ばれました。 aamt.info/news/nagao-stu… 提案法Subset kNN-MTは、長尾先生の用例翻訳の考え方を現代のNMTに融合させた手法であり、大変感慨深く、また、光栄に思います。

@.MBR lovers This looks interesting.

I'll present my poster paper at #CVPR2025 on June 15! I propose an extremely fast post-processing module for diverse nearest-neighbor searches🚀 Y. Matsui, "LotusFilter: Fast Diverse Nearest Neighbor Search via a Learned Cutoff Table" arxiv.org/abs/2506.04790

I start local LLM with #EVOX2

Our long paper has been accepted at #EMNLP2025 main. We present a novel quality-aware decoding, Case-Based Decision-Theoretic decoding (CBDT): 🔥no longer requires pseudo-refs 🚀runs in O(H) time (~100ms) regardless of utility funcs 📈outperforms MBR on MT and image captioning

NICT Sumita-san...! youtube.com/watch?v=HU0fsi…