Language and Learning
GPT-3 and its Few-shot Learning Capabilities – Talk by Prof. Yoav Goldberg
zoomTitle: GPT-3 and its Few-shot Learning Capabilities Abstract: A recent paper by OpenAI, "Language Models are Few-Shot Learners", introduced GPT-3, a huge pre-trained language model, and a claim that it acts as a very effective few-shot learner. While I was skeptic at first, I got access to the GPT-3 model, and after playing with it ... Read more
Diffusion Models Beat GANs? – Eliya Nachmani (TAU & FAIR)
Recording: https://us02web.zoom.us/rec/share/AJjsXIl29nyD0ZSbadWHFd54QJFAB8sfV0b1WdcSY-kMvJE68rmdk595OO_yCUoK.f4P_M3-T4nfR5h91 Eliya Nachmani from TAU & FAIR. Title: Diffusion Models Beat GANs? Abstract: I will give a short introduction to generative diffusion models and present their performance compared to other generative models (including GANs). Moreover, I will present two works that we did in the last year: (i) Noise Estimation for Generative Diffusion Models: ... Read more
Marginal Contribution Feature Importance – an Axiomatic Approach for Explaining Data. Amnon Catav (TAU).
The recording of Amnon's talk: https://us02web.zoom.us/rec/share/TxC7tgud-XyP-h577Pa-EwhLt9Bi4P7BCurUbafwRuW_-UiCxJDG8L9gAJylLiU4.OMMHDUVzuEc4ZVnE?startTime=1636279531000 Title: Marginal Contribution Feature Importance - an Axiomatic Approach for Explaining Data Abstract: In recent years, methods were proposed for assigning feature importance scores to measure the contribution of individual features. While in some cases the goal is to understand a specific model, in many cases the goal is ... Read more
NLP session at AI Week – featuring talks by Prof. Yoav Goldberg and Prof. Ido Dagan
Keynote Yoav Goldberg , Research Director of AI2 Israel; Professor at the Computer Science Department, Bar-Ilan University 11:05 - 11:25 Calendar 1 Calendar Beyond End-to-End: the Case of Multi-text Summarization Ido Dagan , Professor at the Department of Computer Science, Bar-Ilan University; Director, Bar-Ilan Data Science Institute
Investigating fine-tuning of pre-trained language models. Marius Mosbach, from Saarland University
Building 105, Room 106On the 21st of June Marius Mosbach, from Saarland University will give a talk at the lab. He's here for a visit of about 2.5 weeks (starting from June 7th), so be nice to him and say hi =) If you'd like to chat with him (and you should!), add your name in the attached ... Read more
Genomic language model for function prediction of microbial genes – Danielle Miller (TAU)
בחדר 212 בניין 212דוקטורנטית מאת״א תבוא להרצות על מודלי שפה לגנומים בקטריאלים. מוזמנים להגיע ולהפיץ לסטודנטים ולחברי סגל נוספים. ההרצאה תתקיים בשעה 14:00 ביום ב׳ 12.12 בחדר 212 בניין 212. כמו כן, אני מתכנן לשדר אותה בזום: https://us02web.zoom.us/j/5509052372 Title: Genomic language model for function prediction of microbial genes Abstract: Revealing the function of uncharacterized genes is a fundamental challenge in an era of ever-increasing ... Read more
Advanced Machine Learning Methods to Accelerate Materials Discovery (Santiago Miret, Intel)
Building 605, room 11Hi everyone, You are invited to attend the talk of Santiago Miret at the NLP group seminar that will be held on Thursday 9.3.23, at 11 am. The talk will last approximately 1.5 hours including Q&A. See more details below. Best Regards, Idan Location: Building 605, room 11 Title: Advanced Machine Learning Methods to Accelerate Materials ... Read more