there’s something clearly different and more intelligent in a new way, something far more tangible and generally useful than a superb math or coding model
h/t @dorialexander.bsky.social
there’s something clearly different and more intelligent in a new way, something far more tangible and generally useful than a superb math or coding model
h/t @dorialexander.bsky.social
www.inquirer.com/columnists/p...
www.inquirer.com/columnists/g...
www.inquirer.com/columnists/p...
www.inquirer.com/columnists/g...
1. continual learning is a huge focus
2. neolabs are popping up and function to diversify risk in research
1. continual learning is a huge focus
2. neolabs are popping up and function to diversify risk in research
www.theguardian.com/environment/...
Class Incremental Learning from First Principles: A Review
Neil Ashtekar, Jingxi Zhu, Vasant G Honavar
https://openreview.net/forum?id=sZdtTJInUg
#continual #incremental #forgetting
Class Incremental Learning from First Principles: A Review
Neil Ashtekar, Jingxi Zhu, Vasant G Honavar
https://openreview.net/forum?id=sZdtTJInUg
#continual #incremental #forgetting
U n l e a r n i n g
~ in continual scenarios ~
together with:
Zhenyi Wang, Sijia Liu, and Kairan Zhao
6 PM CET / 9 AM PT / 12 PM ET
Hope to see you there!!!
Registration here: stanford.zoom.us/meeting/regi...
U n l e a r n i n g
~ in continual scenarios ~
together with:
Zhenyi Wang, Sijia Liu, and Kairan Zhao
6 PM CET / 9 AM PT / 12 PM ET
Hope to see you there!!!
Registration here: stanford.zoom.us/meeting/regi...
4. researchers are largely in line with Sutton, Karpathy, Ilya — current LLM architecture isn’t enough
tpot gave Dwarkesh a lot of shit for those interviews, but those interviews are right at the pulse of what’s actually happening
4. researchers are largely in line with Sutton, Karpathy, Ilya — current LLM architecture isn’t enough
tpot gave Dwarkesh a lot of shit for those interviews, but those interviews are right at the pulse of what’s actually happening
Neil Ashtekar, Jingxi Zhu, Vasant G Honavar
Action editor: Erin Grant
https://openreview.net/forum?id=sZdtTJInUg
#continual #incremental #forgetting
Neil Ashtekar, Jingxi Zhu, Vasant G Honavar
Action editor: Erin Grant
https://openreview.net/forum?id=sZdtTJInUg
#continual #incremental #forgetting
Some people joke that they have OCD, like it's a cute little foible that is helpful in some way.
OCD is NO FUN, it's not helpful, it can be extremely disabling. It
can be all consuming, it can be hell on earth.
OCD IS NO JOKE.
Some people joke that they have OCD, like it's a cute little foible that is helpful in some way.
OCD is NO FUN, it's not helpful, it can be extremely disabling. It
can be all consuming, it can be hell on earth.
OCD IS NO JOKE.
I'm 4th generation native born, still have relatives there and you continue to piss me off with this continual ignorance to accept science
www.kwch.com/2026/02/18/w...
I'm 4th generation native born, still have relatives there and you continue to piss me off with this continual ignorance to accept science
www.kwch.com/2026/02/18/w...
🎙️ Challenges of Single-Life Continual RL
🎥 www.youtube.com/watch?v=G-Lg...
🎙️ Challenges of Single-Life Continual RL
🎥 www.youtube.com/watch?v=G-Lg...
www.biorxiv.org/content/10.1...
Definitely needed as we hit thousands of eukaryote genomes to compare - shame it won't be *that* easy to run on AMD GPUs (support surely upcoming)
www.biorxiv.org/content/10.1...
Definitely needed as we hit thousands of eukaryote genomes to compare - shame it won't be *that* easy to run on AMD GPUs (support surely upcoming)
Join Rupam Mahmood on Aug 5th @ 11am ET for: “Ways RL Is Failing to Be Continual.” 🧠🤖
🔗 Details & Zoom: lifelong-ml.cc/seminar
✉️ Join our mailing list: groups.google.com/g/collas-list
Join Rupam Mahmood on Aug 5th @ 11am ET for: “Ways RL Is Failing to Be Continual.” 🧠🤖
🔗 Details & Zoom: lifelong-ml.cc/seminar
✉️ Join our mailing list: groups.google.com/g/collas-list
www.youtube.com/watch?v=g-Uf...
#Cancon 🇨🇦 #Mellow #Blues #Jazz
www.youtube.com/watch?v=g-Uf...
#Cancon 🇨🇦 #Mellow #Blues #Jazz