· ai
Ai2's new Olmo 3.1 extends reinforcement learning training for stronger reasoning benchmarks
The Allen Institute for AI Ai2 recently released what it calls its most powerful family of models yet, Olmo 3. But the company kept iterating on the models, exp...