New research from MIT found that those who used ChatGPT can’t remember any of the content of their essays.
Key takeaway: the product doesn’t suffer, but the process does. And when it comes to essays, the process *is* how they learn.
arxiv.org/pdf/2506.088...
18.06.2025 07:32
👍 3718
🔁 1519
💬 192
📌 372
Transformers (how LLMs work) explained visually | DL5
YouTube video by 3Blue1Brown
Finding this a year too late, but an absolutely fantastic visual intro to what's under the hood in transformers by @3blue1brown.com!
www.youtube.com/watch?v=wjZo...
06.06.2025 14:49
👍 0
🔁 0
💬 0
📌 0
Will follow up with a paper thread soon. Do reach out if you have any questions or comments! Excited to hear your thoughts :)
07.05.2025 02:14
👍 0
🔁 0
💬 0
📌 0
A neurally constrained computational model of context-dependent fear extinction recall and relapse - Communications Biology
A neurally-constrained, computational model of cued fear conditioning proposes potential mechanisms underlying the context-dependent relapse of fear following fear extinction, and suggests that counte...
Excited to share my first-author paper with Dr. Thad Polk in @commsbio.nature.com introducing a neurally constrained, memory engram-based mechanistic model of the fear circuit to explain context-dependent relapse of extinguished fear(ConFER), and present clinically relevant hypotheses! #compneuro
07.05.2025 02:14
👍 6
🔁 2
💬 1
📌 0
Thank you! 😄
06.05.2025 17:10
👍 0
🔁 0
💬 1
📌 0
Recently defended my dissertation titled 'A Neurally Constrained Computational Model of Context-Dependent Fear Extinction and Relapse' at @umich.edu with Dr. Thad Polk as chair, and an incredible committee comprising of @natalie-t.bsky.social, Dr. Richard Lewis and Dr. Elizabeth Duval!
05.05.2025 20:42
👍 20
🔁 0
💬 1
📌 0
Congrats Kali!! 👏
05.05.2025 19:18
👍 1
🔁 0
💬 1
📌 0