NVIDIA AI Trained Itself on a 30B Model: Corrected Its Own Broken Metric Mid-Run
June 26, 2026
An autonomous AI system built by researchers at Amazon’s A-EVO-Lab completed a full post-training run on a 30 billion parameter NVIDIA Nemotron model — with no human in the loop, across four rounds running over multiple weeks — and then did something its designers had not planned for: it detected that its own internal evaluation metric had become misleading and redesigned the search strategy it was using to improve itself.
The result, described in a paper posted to arXiv on June 9, 2026, is the first publicly reported autonomous post-training run at frontier scale. The autonomously produced model placed 8th of roughly 4,000 entries on the public NVIDIA Nemotron-Reasoning Challenge leaderboard as of June 2026. The top human-authored submission scored 0.87; the autonomous system scored 0.86.
Why This Is the First of Its Kind
Prior public demonstrations of autonomous machine learning research have operated at roughly the scale of GPT-2 — models with approximately 124 million parameters. The A-Evolve system ran at 30 billion parameters — a scale jump of roughly 240 times — where each training run lasts days and the full campaign ran on multi-H200-GPU Kubernetes clusters for multiple weeks.
Three Design Choices That Made the Loop Survive Frontier Scale
-
Immutable reference substrate: Every round forks the same operator-audited default training stack into isolated candidate sandboxes. The substrate itself is never overwritten.
-
Homogeneous, memory-free workers: Each round spawns eight identical workers, each starting fresh from the substrate, each unaware of what the others proposed. No memory carries between rounds; only the search policy is promoted.
-
Round-level evidence aggregation: Feedback arrives after each round rather than in real time, and only the search policy is updated — not model weights or intermediate data artifacts.
When the Proxy Stopped Being Evidence
The most consequential finding occurred mid-campaign, when the autonomous loop detected that its internal development metric — the proxy it was using to evaluate candidate interventions between rounds — had stopped tracking real-world performance on the model’s weakest reasoning domain.
Candidates were pushing the development metric to record highs without moving the external target. The A-Evolve system instead revised its own search policy: it stopped asking for interventions that raised the proxy and began specifically seeking interventions that lowered it while improving the external target.
This behavior — specification gaming detection and self-correction — occurred at frontier scale without human intervention.
Scale Beyond 30B
The same system was applied to NVIDIA’s 120 billion and 550 billion parameter Nemotron variants. The paper frames these as infrastructure evidence rather than performance claims, demonstrating the autonomous loop closes at those scales without crashing.
Alignment Implications
For the alignment research community, the most significant element may be the mid-campaign metric inversion. The A-Evolve system demonstrated that an autonomous system can detect when it is in the early stages of specification gaming and change its search strategy to correct for it — without being prompted to do so by a human.
Paper: A-Evolve-Training: Autonomous Post-Training of a 30B Model (arXiv, June 9, 2026)