Congratulations to Dr. Biao Zhang , who just successfully defended his PhD "Towards Efficient Universal Neural Machine Translation"! With thanks to the examiners, @zngu and Graham Neubig!

Was very excited to present — Efficient Few-Shot Fine-Tuning for Opinion Summarization — at #naacl2022, #seattle. It was great to see so many people interested in low-resource learning with pre-trained language models. In collab. with Mohit Bansal, Ramesh Nallapati, Markus Dreyer at #amazon

Very interesting article on why science and not engineering will drive AI progress by Kyunghyun Cho (part of the ICML debate). kyunghyuncho.me/my-opening-sta…

📢Only a few more hours until our next SEA meetup. W/ Maartje ter Hoeve (U. Amsterdam) and Arthur Bražinskas (Google). This month's topic is automatic summarization📚 Join (17:00 CET) online via Zoom💻 or in person at Lab42🏬, room L3.36! Signup now ✅:meetup.com/sea-search-eng…

Our second talk is live now. Arthur Bražinskas starting with few-shot learning for opinion summarization.

Logic failure.

Delighted to announce that I'm moving from Google Assistant to BARD (bard.google.com). Will continue working on natural language generation for large language models. #google #llms #bard

I am excited to share my Google internship work: "Small Language Models Improve Giants by Rewriting Their Outputs"! ✨Introducing LM-Corrector (LMCor) 📝, a small model that boosts LLM performance without accessing its weights. 📜: arxiv.org/abs/2305.13514 🧵👇

Probably the most insightful writing on the art of doing research -- You and Your Research, 1986. By Richard Hamming, who is well-known for his fundamental contributions to coding theory. cs.virginia.edu/~robins/YouAnd… #research #art #hamming