AI RESEARCH
Jupiter-N Technical Report
arXiv CS.CL
•
ArXi:2604.17429v1 Announce Type: new We present Jupiter-N, a hybrid reasoning model post-trained from Nemotron 3 Super, a fully open-source 120B parameters LLM. We target three objectives: (1) agentic capability via uncertainty-curated trajectories; (2) UK cultural alignment via synthetic data grounded in cultural norms; and (3) Welsh language via parallel corpora and LLM-translated Welsh conversations.