reinforcement learning
-
Poolside models and the GitHub Copilot transition
Paris-based startup Poolside approaches a $3 billion valuation with its malibu and point foundation models. These models utilize reinforcement learning from code…
-
The evolution of DeepMind and the progression of Go software
DeepMind’s technical progression from AlphaGo Fan to AlphaZero reshaped global AI research funding. Following its Google acquisition, AlphaGo Lee defeated Lee Sedol…
-
The evolution of DeepMind
Demis Hassabis and his co-founders established DeepMind in 2010 to develop general-purpose AI. The company’s 2016 AlphaGo victory over Lee Sedol demonstrated…
-
Sakana AI optimizes model performance through evolutionary merging
Sakana AI utilizes evolutionary techniques to combine open-source models, creating specialized systems like the EvoLLM-JP-v1-7B which outperforms larger 70B parameter models. The…
-
How o1 Changes the LLM Training Picture, Part 2: Search, Reward and Cost
What reasoning models borrow from game-playing systems, why the reward signal is the bottleneck outside maths and code, and when the extra…