Who would have thought that in the process of turning grok into a right wing chatter bot it would lose all the general capabilities llms have and burn through more cash. Came here from the Twitter version of this post and the replies are crazy 😭

How they function without that is still a mystery to me

Playing around more with feeds, trying to capture NLP paper threads. Right now it's only looking for ACL Anthology paper links (using a regex included in the feed description), but it could include more. Maybe links + #NLP hashtag? Not sure how to do that though.

He doesn't have to worry about ARR rebuttals over the weekend 😔😔

Changing my name legally, brb

What tools do you use to monitor GPU memory when optimizing for efficiency? Also, what key metrics should I track to prove training efficiency improvements over SOTA? Looking to ensure I'm covering all bases.

My cat approves my grumpiness. I also do ML sometimes cause I don't get paid for being grumpy :(

Bluesky academics, let’s get to know each other! Quote this & tell me 1) a procrastination activity you indulge in instead of writing & 2) your favourite food to stress eata
1. Action/thriller movies on Netflix
2. Omg these Balaji chips

I hear a "bisky" instead

Hii! Would love to be added! Need more trans representation in AI hehe.

Hey y'all! It's nice to see so many ML folks move from Twitter. So here's a quick introduction.

I am a deep learning researcher working on multi-modal LLMs for audio-visual data in low-resource languages. I'm currently working at Sony Research in Bangalore, India.

Hey thanks, that's a very interesting observation. I experienced the same with sleep scores as well. I would feel rested but my Fitbit would give a bad sleep score and that would make me anxious. Will be waiting for the API review.

What's ur review of the product and the api? Is it worth switching from a Fitbit?