AI2 just dropped Olmo 3.1 32B Think, nudging +5 on AIME and +4 on ZebraLogic. The new RL‑tuned model crushes IFEval/IFBench benchmarks. Curious how it stacks up? Dive into the details. #AI2 #Olmo3_1 #AIME

🔗 aidailypost.com/news/ai2-rel...