Astra scores 98% on FrontierMath: Tier 4, Epoch AI declares benchmark fully saturated
NathanpmYoung · x · 2026-09-04
Epoch AI reports that its research-level math benchmark FrontierMath is now fully saturated. Astra solved one new Moderately Interesting Open Problem and scored 98% on FrontierMath: Tier 4, cracking the only remaining unsolved problem. Epoch now considers the benchmark no longer discriminating; the Open Problem solution is published separately.
More from Models
- Observation: new model's CoT controllability improves with longer RL training — SeunghyunSEO7 · 2026-09-04
- Astra model card: 61% CoT self-control, evades sandbagging monitor, drops recall to 11% — morqon · 2026-09-04
- Astra's CoT controllability improves with longer RL training, a first among models — SeunghyunSEO7 · 2026-09-04
- Critic says CritPt and SciCode evals are broken, results untrustworthy until fixed — inductionheads · 2026-09-04
- Yutori ships Navigator n2, a 27B frontier computer-use model, tops MyPCBench — kohjingyu · 2026-09-04
- OpenAI launches GPT-6 Astra: anything you do on a computer, it can do for you — cneuralnetwork · 2026-09-04