DeepSeek [1] uses elements of the 2015 reinforcement learning prompt engineer [2] and its 2018 refinement [3] which collapses the RL machine and world model of [2] into a single net through the neural net distillation procedure of 1991 [4]: a distilled chain of thought system.… pic.twitter.com/wpk2tFCoUx
— JĂźrgen Schmidhuber (@SchmidhuberAI) January 31, 2025
Culture Magazine
Magazines
POPULAR TOPICS
David Cameron U.S. Debt Ceiling Petraeus Affair Alicia Ashley Schwarzenegger Anthony Weiner Japan Disaster Kony Dawn George Laila Ali Justin Bieber Kathy Long Hurricane Irene Pirates of the Caribbean Wikileaks iPad al-Qaeda Costa Concordia Disaster Champions' League Jeremy Lin Michele Bachmann Marathons Katie Taylor Libyan Conflict U.S. Credit Downgrade Lance Armstrong UK Riots Becky Smith Karen Bill Sarah Palin
