I just mean long term, imo RL finetuning paradigm is a big upgrade over just SFT (expert imitation) for LLMs at the current stage of development and will continue to grow substantially.
RESEARCH
-
Humans Enrich Animal Behavior Through Cultural and Technological Lens
By
–
Humans take every animal behavior and make it a million times richer.
-
Training Data Evolution: Internet Text to Supervised Conversations
By
–
In era of pretraining, what mattered was internet text. You'd primarily want a large, diverse, high quality collection of internet documents to learn from.
— Andrej Karpathy (@karpathy) 27 août 2025
In era of supervised finetuning, it was conversations. Contract workers are hired to create answers for questions, a bit… https://t.co/rR6yYZGgKPIn era of pretraining, what mattered was internet text. You'd primarily want a large, diverse, high quality collection of internet documents to learn from. In era of supervised finetuning, it was conversations. Contract workers are hired to create answers for questions, a bit
-
Von Neumann’s Achievements: Game Theory, Computing, Nuclear Weapons
By
–
Einstein wasted the second half of his life on a fruitless quest. In the second half of his life, von Neumann invented game theory, computer architecture, implosion nuclear weapons, cellular automata and weather prediction, among other things.
-

Code vs Data: The Future of Robotics Debate
By
–
Will the future of robotics be written in code or in data? A debate at ICRA among MIT profs. & their international colleagues revealed that we'll probably need both to unlock robots' real-world potential — but which takes the lead, & when, isn't clear yet:
-
Aleph Retains Complex Scenes Without Masking Requirements
By
–
Aleph can retain complex scenes and fine details without the need for tedious masking. pic.twitter.com/pCh3avPKkA
— Runway (@runwayml) 27 août 2025Aleph can retain complex scenes and fine details without the need for tedious masking.
-

Where to Buy Adversarial Examples AI Sweatshirt Merchandise
By
–
how can i buy an Adversarial Examples sweatshirt? (for real. i would wear it)
-
NVIDIA Graduate Fellowship Awards: $60,000 Applications Open
By
–
Applications open through September 15 for $60,000 NVIDIA Graduate Fellowship Awards!
-

Google DeepMind AI for Organizations Grand Challenge Launch
By
–
HAI and @GoogleDeepMind invite scholars to submit bold ideas on how AI can improve and reimagine organizations. Join our virtual Q&A tomorrow to learn more about the AI for Organizations Grand Challenge: 9 am PT: http://
bit.ly/45DiheH 6 pm PT: https://
bit.ly/3HwsxfD -

Nvidia Presents Autoregressive Universal Video Segmentation Model
By
–
Nvidia presents Autoregressive Universal Segmentation Model