{"version":1,"type":"story","url":"https://digestai.news/story/nemotron-3-ultra-pipeline-achieves-imo-gold-with-open-weight-release","json":"https://digestai.news/story/nemotron-3-ultra-pipeline-achieves-imo-gold-with-open-weight-release.json","markdown":"https://digestai.news/story/nemotron-3-ultra-pipeline-achieves-imo-gold-with-open-weight-release.md","slug":"nemotron-3-ultra-pipeline-achieves-imo-gold-with-open-weight-release","headline":"Nemotron 3 Ultra pipeline achieves IMO Gold with open-weight release","summary":"NVIDIA has released an open-source framework that enables its Nemotron 3 Ultra model to solve Olympiad-level mathematics problems at a gold-medal standard. The system utilizes a test-time compute pipeline that iteratively generates, verifies, and refines natural-language proofs without relying on formal provers or external tools.\n\nThe approach involves training two specialist checkpoints from the base Nemotron 3 Ultra model using supervised fine-tuning and reinforcement learning. By combining these specialists with the general-availability model in an iterative search process, the system achieved a score of 30 out of 42 points at the International Mathematical Olympiad (IMO) 2026, surpassing the threshold for a gold medal.\n\nTo support further research, the team has released the post-trained checkpoints, training data, inference code, and submitted solutions. They also introduced Nemotron-IMO-Bench, a new benchmark consisting of 200 novel olympiad-level problems, providing the community with resources to evaluate and improve mathematical reasoning capabilities in large language models.","keyPoints":["Nemotron 3 Ultra pipeline scored 30/42 at IMO 2026, meeting the gold-medal threshold.","System uses iterative natural-language proof generation, verification, and refinement without formal tools.","NVIDIA released two specialist checkpoints, training data, code, and a new 200-problem benchmark."],"whyItMatters":"This demonstrates that open-weight models can achieve frontier-level performance in complex mathematical reasoning, lowering the barrier for researchers to develop and evaluate advanced AI systems for rigorous problem-solving tasks.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["NVIDIA"],"models":["Nemotron 3 Ultra","Nemotron-IMO-Bench"],"people":[]},"firstPublishedAt":"2026-09-12T04:00:00Z","updatedAt":"2026-09-12T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics","url":"https://arxiv.org/abs/2609.10712","publishedAt":"2026-09-12T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Nemotron 3 Ultra pipeline achieves IMO Gold with open-weight release\", 12 September 2026, https://digestai.news/story/nemotron-3-ultra-pipeline-achieves-imo-gold-with-open-weight-release","publisher":"Digest AI","title":"Nemotron 3 Ultra pipeline achieves IMO Gold with open-weight release","datePublished":"2026-09-12T04:00:00Z","url":"https://digestai.news/story/nemotron-3-ultra-pipeline-achieves-imo-gold-with-open-weight-release"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}