An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics
Summary
The paper presents an open-method using post-trained Nemotron 3 Ultra checkpoints to achieve gold-medal performance on IMO 2026 through iterative verification and refinement in natural language without external tools.
View Cached Full Text
Cached at: 09/11/26, 06:16 AM
Paper page - An Open Recipe for IMO Gold: Training Nemotron for Olympiad Mathematics
Source: https://huggingface.co/papers/2609.10712
Abstract
A natural-language proof-generation pipeline using post-trained Nemotron 3 Ultra checkpoints achieves gold-medal performance on IMO 2026 through iterative verification and refinement without external tools.
We study how model post-training and test-time inference design affect natural-language proof generation for hard olympiad mathematics. Starting from Nemotron 3 Ultra, we train two specialist checkpoints usingsupervised fine-tuningandreinforcement learning, and evaluate checkpoint choice,verification, andrefinement. Based on these findings, we present an open-model test-time-compute pipeline. The system operates entirely in natural language, with no formal prover, external tools, or internet access. Three Nemotron 3 Ultra checkpoints - the general-availability model and two post-trained specialists - power aniterative searchthat generates, verifies, and refines candidate proofs; a separate high-compute stage then selects each final submission. The system scored 30 out of 42 points at IMO 2026, reaching the gold-medal threshold. We release the two post-trained checkpoints as well as the training data, the training and inference code, the submitted solutions, andNemotron-IMO-Bench, a new benchmark of 200 novel olympiad-level problems.
View arXiv pageView PDFAdd to collection
Get this paper in your agent:
hf papers read 2609\.10712
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2609.10712 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2609.10712 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2609.10712 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO
NVIDIA 展示了如何基于 Nemotron 3 通过 SFT、RL 和 GenCorrect 推理循环进行专业化微调,在 IOI 2026 和 IMO 2026 上同时达到金牌水平,证明了一个可复用的专业化训练配方。
@ClementDelangue: Paper of the day! https://huggingface.co/papers/2605.13301…
A paper introduces a unified recipe (SU-01) that combines reverse-perplexity curriculum, two-stage reinforcement learning, and test-time scaling to achieve gold-medal-level performance on IMO and IPhO problems using a 30B-A3B backbone.
Achieving Gold-Medal-Level Olympiad Reasoning via Simple and Unified Scaling
This paper presents a simple and unified recipe combining supervised fine-tuning, two-stage reinforcement learning, and test-time scaling to train a reasoning model (SU-01) that achieves gold-medal-level performance on International Mathematical and Physics Olympiad problems.
Achieving Gold-Medal-Level Olympiad Reasoning via Simple and Unified Scaling
A paper presenting SU-01, a 30B-A3B reasoning model that achieves gold-medal-level performance on IMO and IPhO problems via reverse-perplexity curriculum, two-stage reinforcement learning, and test-time scaling.
Maxproof
MaxProof introduces a test-time scaling framework that combines proof generation, verification, and repair using generative-verifier RL, enabling the M3 model to exceed human gold-medal thresholds on IMO 2025 and USAMO 2026.