UPDATE:
I sent my first PhD application on October 30th.
21 applications, and 6 stressful months later, I've decided to join @MPI_EvolBio for my PhD.
I will soon graduate with a BS-MS (in Physics!) from the Indian Institute of Science, Bangalore.
Nine PhD positions are open at Max Planck Institute. Come to work with me on modeling of the origins of organismic complexity at the onset of multicellular life. Application deadline is January 26.
Prof Rohini Godbole was much more than a renowned scientist. She was a leader, and a strong advocate of women in science. She was a friend who will be missed. https://t.co/AOyXkQXJy4
I couldn't resist. Since this week seems to be a lot about the boundaries between AI and physics, I couldn't help but notice that economic complexity also seems to fit the mold :-)
@hinduhuman1997@vishuguttal Coincidentally, I put out an interactive explainer on Hopfield networks just a month ago. I hope this link answers your questions
https://t.co/jUwOZxGZTK
@hinduhuman1997@vishuguttal Coincidentally, I put out an interactive explainer on Hopfield networks just a month ago. I hope this link answers your questions
https://t.co/jUwOZxGZTK
Despite being a very simple framework, Hopfield networks are capable of learning and recalling patterns. They also feature associative memory, in the sense that they can associate partial/noisy patterns with the learned ones.
Interactive visualization: https://t.co/E0UDr5tbhd
Surprised and delighted to notice that my website, https://t.co/GZa5YhToC4, has been cited by a paper about using ML to augment static math textbooks, published in an ACM journal: https://t.co/oOFZbprFbc
📢📢 I will be on PostDoc job market from Summer 2025!!
Grateful for RTs 😇& my CV is here: https://t.co/AmrW5pVh7R
If you know someone who is hiring and is a good fit, please forward it :)
I will also be at #FOCS2024 next month. DM if you are around, would love to catch up👋
Despite being a very simple framework, Hopfield networks are capable of learning and recalling patterns. They also feature associative memory, in the sense that they can associate partial/noisy patterns with the learned ones.
Interactive visualization: https://t.co/E0UDr5tbhd
Despite their limitations, Hopfield networks can provide insights about how LLMs may store and retrieve information, and they also try to (simplistically) replicate how the brain probably uses snippets of information to recall large segments of memory from a humungous repository.
Despite being a very simple framework, Hopfield networks are capable of learning and recalling patterns. They also feature associative memory, in the sense that they can associate partial/noisy patterns with the learned ones.
Interactive visualization: https://t.co/E0UDr5tbhd
The recall phase with a partial/noisy pattern involves tweaking the nodes such that the pattern descends towards a local minima.
Thanks to this video by @ArtemKRSV which explains all the math: https://t.co/XsSrYgR2qm
This is cool, good for the OSS devs.
Hey! @_chanrt_ you should try applying for this @Sourcegraph@SourcegraphCody for your repo - https://t.co/6uNXI8KUZQ and you will get all the top LLMs for you development purposes in your IDE with Cody extension.