Atlas · C09 · Core cognition
Quantitative & mathematical
Working with numbers, quantities, statistics and mathematical reasoning.
AI Reliable score
8 / 10
Expert level · peak 10 · trend ↑ · grade A
Why this score
Gold-medal standard at the 2025 International Mathematical Olympiad; near-perfect on AIME; 31% on FrontierMath Tier 4 (research-level). Slips in long calculations without tools keep Reliable at 8.
On the human scale
2Beginner. Basic arithmetic
4Average adult. Average adult: percentages, simple algebra
6Competent pro. Professional: applied maths and statistics for work
8Expert. Expert: advanced proofs, research-grade applied maths ← AI today
10Best humans. Top mathematicians
Future warning
At its best, AI already reaches 10 on this capability. Peak results are never counted in the index; they show where the water may rise next.
Evidence
IMO 2025 gold-medal standard (Google DeepMind) + FrontierMath Tier 4 (Epoch AI) + AIMEA
E12 · 2025-07 / 2026-01
- Measures
- Olympiad and research-level mathematics
- AI result
- Gemini Deep Think officially gold-medal standard at IMO 2025; FrontierMath Tier 4 record 31% (GPT-5.2 Pro, Jan 2026); AIME 2025 99%.
- Humans
- IMO gold = top ~8% of the world's best high-school mathematicians
- On our scale
- Expert to virtuoso peak.
deepmind.google ↗
epochai.substack.com ↗
A Definition of AGI (Hendrycks et al., CHC-based)B
E07 · 2025-10
- Measures
- 10 CHC cognitive domains, each worth 10% of an 'AGI score'
- AI result
- GPT-5 total 57%: Knowledge 9, Reading-Writing 10, Maths 10, Reasoning 7, Working memory 4, LT memory storage 0, LT memory retrieval 4, Visual 4, Auditory 6, Speed 3 (out of 10 each).
- Humans
- Well-educated adult (definition of 100%)
- On our scale
- Used as a cross-check on the profile shape, not for absolute levels (single model, dated Oct 2025).
arxiv.org ↗
History
First reading: Atlas v0.1, 5 Nov 2026. Every later change will be listed here with its reason.
Disagree? Send a correction
