Will be at #emnlp2023 as a co-organizer of the BLP and CALCS workshops. Looking forward to present the MultiCoNER v2 paper, a dataset on fine-grained complex NER in 12 languages. Feel free to say hi. Will be happy to talk about research and music!

ICLR has placed OpenReview in a difficult position, so I want to offer a few words about the OpenReview team working behind the scenes. OpenReview has long been operated at UMass Amherst as a non-profit organization founded by Andrew McCallum. Each year, Andrew must raise more

This is a great scope to actually make impact on the system rather than just talking on the social media :)

The submission deadline for the 2nd Workshop on Bangla Language Processing (BLP) workshop (co-located with AACL-IJCNLP 2025) has been extended to October 4, 2025 ( https://t.co/rcWuSfizmv) @firojalam04 @cryptexcode @Enamul_Hoque @shammur_absar @MdNishatRaihan @aaclmeeting

Can we say that #Coldplay live production team hallucinated just like our LLMs do?

I am kinda enjoying the LLM powered auto-complete plugins for Visual Code throughout the last couple of years. Suggestions are mostly okay. But what irritates me is the UX. Sometimes I type tab, because I want an indent. And what I get is 15 lines of code vomit to clean up. Wby?

Early days of Google Translate were hilarious. But this translation approach with unnecessary reasoning and eventually poor translation by DeepSeek R1 (8B and 32B) is whole different level, ROFL.

I created a Python project starter repo for students that helps maintain good code quality while doing research projects: https://t.co/HRFdxAucsI I was opinionated and made only one choice for each tool, but there are other options too!

📣 Did you miss the @aclmeeting deadline? No worries! We're excited to announce that the paper submission deadline has been extended to February 21. If you have papers related to code-switching, submit them here. Don't forget to participate in our shared task as well!

Don't forget to submit your paper to CALCS!

📣 We are excited to host CALCS at this year's @naaclmeeting and to announce our call for papers. For more details, please visit https://t.co/CY527XCdx2. This year, we are also introducing a new code-switching shared task. Stay tuned for more details! #NLProc #naacl2025

Very anxious about friends and in general every person and their families in Los Angeles. Hope they will be safe and the situation will get better. Climate change is real and needs more attention than money making machines.

2025 will be the first year in the history with strong model capabilities getting integrated into consumer grade products at scale. Happy new year!

2024 is very memorable for me. Personally, became a dad. Professionally, I had to spend a ridiculous amount of time on social media to stay updated with new models and capabilities popping up here and there.

Modern LLMs are designed to "please" consumers through tone, style, content. When it comes to LLMs use in education, some stages in learning requires not-so-pleasant-to-hear and often harsh feedback, strictness, and truth-on-face. How LLMs would be designed to handle this?

~6 years ago, I read #Sapiens after 2 psychology books - Take Pride and The Lucifer Effect. The combination of these three probably made me quite pessimistic as a person. Now, #Nexus is on my desk for a week, and I am scared to start reading it. 🤷‍♂️

What's your favorite LLM backed product? Mine is NotebookLM.

Was going through old files from old laptop and found a folder of scratch codes from 2016-17. Was implementing Attention using Keras and Pytorch, testing with made up vectors with LSTMs. Life was beautiful back in grad school.

Cheers to everyone helping with the review process. Double cheers to the great reviewers ...

🚨 We are hiring! 🚨 The ACL Rolling Review is looking for motivated individuals to join our team! If you're passionate about NLP and want to contribute to the community, apply now! Editorial Assistant - https://t.co/caJIIA2gGl Workflow Manager -

The spirit of open-source is straightforward. Anyone should be able to break it down, investigate, reproduce, improve, fork and turn into something way different. If your release is not enabling it, stop labeling it as open source.