El editorial de EL PAÍS España advierte sobre el impacto ambiental de la IA. Es crucial hacer modelos más eficientes. Pero la prensa insiste en ver la IA como amenaza. Mientras, ninguna empresa española entrena grandes modelos y los beneficios se van fuera elpais.com/opinion/2024-0…

I'm going to present our paper Understanding Reasoning Ability of Language Models From the Perspective of Reasoning Paths Aggregation from 11:30 am to 1 pm tomorrow (7.24) at Hall C 4-9 #812. Stop by our poster if you are interested!

How do LLMs gain their abilities from pre-training data? Generalization vs Memorization Check this out 👇

We (a group of University of California students and faculty from 7 campuses, along with researchers from over 20 other institutions) are releasing an open letter in opposition to California SB-1047. Read the letter here: drive.google.com/file/d/1E2yDGX… In particular, our concerns are:

Our work “MultiAgent Collaboration attack” got accepted to Findings at EMNLP ! 🥳 Check it out and reach out if you’re interested 👇