arrow_back
Threads
Threads 1257
Threads
New notes from the Stanford CS336: Language Modeling from Scratch
Threads
another banger dropped by @jbhuang0604
Threads
I thought I had a good /plan custom Claude Code command.
Threads
if they give this shit away for free it can't be profitable right https://t.co/8seGjNOWcJ
Threads
exactly same feeling. It’s hard to believe current RL is the right way to AGI
Threads
Every single talented / accomplished ml researcher I’ve met so far care *a lot* about making algorit
Threads
Claude Code Model Switcher
Threads
1/ “Wider or Deeper? Scaling LLM Inference-Time Compute with Adaptive Branching Tree Search” will be
Threads
Image to @threejs model? no problem
Threads
I'm reading NVIDIA's new paper and its wild.
Threads
Really interesting paper that fills the gap between GRPO theory and practice.
Threads
trying to be disciplined is the quickest way to fail in life.
Threads
@gautam_next @TheNaveena The Smart Multilevel Rotary Parking System at KBR Park was developed by Nav
Threads
Must read piece by young scholar @Suyash_Desai on China’s effort to military cold start posture. Rel
Threads
The World of Us (dir: Yoon Ga-eun, 2016) https://t.co/T0Ch2GNbrw
Threads
it’s incredible how much better people get at math once you start explicitly mapping the colors out
Threads
Pretrain a BERT Model from Scratch
Threads
Oh, and roughly 30x cheaper than Gemini 3.0 pro
Threads
At neurips and interested in continual learning. Please DM me if you’re interested talking.
Threads
"We need some time to find ourselves."
Threads
DeepSeek V3.2 finally corrected the advantage calculation (in red).
Threads
Advent of Claude Day 2 - Double Esc to Rewind
Threads
These youtube videos are a banger to understand
Threads
https://t.co/HdQHHqj3oO
auto_stories
Select a bookmark to read
Click any item from the list