True Index: Domain Rankings

Private Top-Tier (S & A) standing across 3 core domains
Agentic & Enterprise22.4p
#38
4 Top-Tier private tests
Coding & SWE20.2p
#38
4 Top-Tier private tests
Math & Scientific Reasoning46.9p
#24
2 Top-Tier private tests

Head-to-Head: Clean Evals vs. Legacy Benchmarks

4 matched domains · positive boost means higher legacy scores
Capability Domain Clean Evals (Tiers S & A) Legacy Benchmarks (Tiers C & D) Public Lift (Boost)
Agentic & Enterprise22.4p (4)27.7p (2)+5.3p
Coding & SWE20.2p (4)50.0p (1)+29.8p
General Knowledge & QA37.6p (2)13.9p (1)-23.8p
Math & Scientific Reasoning46.9p (2)28.2p (2)-18.7p
Clean Evals Avg (Tiers S & A)
31.8p
12 private, leak-resistant tests
Legacy Benchmarks Avg (Tiers C & D)
30.0p
6 public or aging tests
Matched Public Lift (Boost)
-10.3p
Near zero or negative indicates clean generalization
Machine-Readable Export

Audited Benchmark Performance of Step 5 Preview

Grouped by benchmark integrity so private capability signal stands apart from aging public tests.

Top-Tier Validated Tests (Tiers S & A) (12)

Private, leak-resistant tests that build real capability signal.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
MLCR-AAPrivate / FreshMath & Scientific Reasoning#28/4643.0p
AA-Briefcase v1.1Private / FreshAgentic & Enterprise#35/5030.6p
AutomationBench-AAPrivate / FreshAgentic & Enterprise#40/5020.4p
MKA (Multimodal Knowledge Agent)Private / FreshAgentic & Enterprise#23/2317.8p
FrontierCode-AA 2.0Private / FreshCoding & SWE#41/4612.0p
Code2Video-AAPrivate / FreshCoding & SWE#22/2311.6p
AA-Omniscience v1.1Private / FreshGeneral Knowledge & QA#24/4751.3p
CritPt v2Private / FreshMath & Scientific Reasoning#18/4050.8p
Code Migration v1.1Private / FreshCoding & SWE#17/2431.9p
AA-3DPrivate / FreshCoding & SWE#21/2225.4p
LiveBenchPrivate / FreshGeneral Knowledge & QA#26/2824.0p
AA-AnalystAgentPrivate / FreshAgentic & Enterprise#22/2420.9p

Mid-Tier Evaluations (Tier B) (1)

Partially exposed or transitional evaluations with moderate leakage risk.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
HLE-DiamondSemi-ExposedGeneral Knowledge & QA#28/4943.8p

Legacy & Vulnerable Benchmarks (Tiers C & D) (6)

Public or aging benchmarks prone to overfitting and training-data leaks.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
Terminal-Bench 4.0Public / AgingCoding & SWE#25/5050.0p
BioMysteryBenchPublic / AgingMath & Scientific Reasoning#18/2336.8p
GDPval-AA v2.1Public / AgingAgentic & Enterprise#27/4335.0p
GDP.pdfPublic / AgingAgentic & Enterprise#40/5020.4p
Terminal-Bench-Science 0.1Public / AgingMath & Scientific Reasoning#35/4119.6p
MMMU-ProPublic / AgingGeneral Knowledge & QA#19/2213.9p