Nemotron 3 Ultra: Open Moe Hybrid Mamba-Transformer for Agentic Reasoning [pdf](research.nvidia.com) |
Nemotron 3 Ultra: Open Moe Hybrid Mamba-Transformer for Agentic Reasoning [pdf](research.nvidia.com) |
It is significantly bigger than Qwen for the same level of intelligence, but I think the key strength was inference speed.