ResearchExplainer
Axion 1.0 trains an agent model with two SFT stages and RL
The open recipe skips mid-training and DPO, and its 122B model scores 59.0 on Humanity's Last Exam. Weights for the 122B model are not clearly released.
Axion
Every ClaudeGPT Si story about Axion, newest first: news, analysis, reviews and guides.
1 story · Updated