A 30B MoE model with only 3B active parameters achieves Gold Medal-level performance in International Math and Informatics Olympiads.
March 20, 2026
Original Paper
Nemotron-Cascade 2: Post-Training LLMs with Cascade RL and Multi-Domain On-Policy Distillation
arXiv · 2603.19220
The Takeaway
It demonstrates extreme 'intelligence density' by matching the reasoning capabilities of frontier models (671B+) with 20x fewer parameters. This suggests that the ceiling for compact model reasoning is much higher than previously thought when using Cascade RL and multi-domain distillation.
From the abstract
We introduce Nemotron-Cascade 2, an open 30B MoE model with 3B activated parameters that delivers best-in-class reasoning and strong agentic capabilities. Despite its compact size, its mathematical and coding reasoning performance approaches that of frontier open models. It is the second open-weight LLM, after DeepSeekV3.2-Speciale-671B-A37B, to achieve Gold Medal-level performance in the 2025 International Mathematical Olympiad (IMO), the International Olympiad in Informatics (IOI), and the ICP