Thank you, Aran, for sharing our work!

Really excited that our paper 'A PhD Student’s🧑‍🎓 Perspective on Research in NLP in the Era of #LLMs ' will be at #COLING2024 ! We brainstormed 45 topics💡 that students can work on despite the LLMs. Co-led by Oana Ignat 👩‍💻🎓🇷🇴 Zhijing Jin and Rada Mihalcea with contributions from many.

Want to know more about #LargeLanguageModels ?

PhD candidate Muhammad Khalifa Muhammad Khalifa explores the challenges #LLMs face with chain-of-thought (multi-step) reasoning, in this talk for ML Collective:
youtube.com/watch?v=x7kg0J…

I enjoyed giving the talk thanks for the invite! Rosanne Liu

📢Join us to learn about Discriminator-Guided Chain-of-Thought Reasoning in DLCT! ✨Excited to have Muhammad Khalifa as our speaker.

Mark your calendars for March 1st at 10 AM PT. Don't miss out!

#DLCT #AI #Reasoning #LLM

More info & upcoming talks: mlcollective.org/dlct/

Enjoying this new feature by arXiv.org where you can view a paper in HTML format!
No further need to clutter my phone with PDFs that I will read only once

screenshot from this paper:
browse.arxiv.org/html/2312.1144…

Cool idea, but I don't fully understand the need to do implicit CoT. There's great value in the intermediate textual thoughts for evaluation, explainability, etc. as they we are the window into the model's reasoning process. I don't think we should compromise that

I'm giving a virtual talk next week at NAVER LABS Europe
Sign up link here: europe.naverlabs.com/research/semin…

News media often select content that supports their ideology, but why do they report events that favor out-group members? We introduce PAC, a dataset with PArtisan and Counter-partisan events.

⭐️Paper: arxiv.org/abs/2310.18768
⭐️Dataset: github.com/launchnlp/Part…

#EMNLP2023
🧵👇

🎯 Tired of one-size-fits-all AI chatter? ChatGPT tends to generate verbose & overly informative responses. This is because the current RLHF pipeline only allows aligning LLMs to the general preferences of the population. However, in the real world, people may have multiple,…