True Index: Domain Rankings

Private Top-Tier (S & A) standing across 3 core domains
Agentic & Enterprise60.2p
#18
2 Top-Tier private tests
Coding & SWE59.4p
#19
1 Top-Tier private test
Math & Scientific Reasoning43.0p
#29
1 Top-Tier private test

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

3 matched domains · positive boost means higher legacy scores
Capability Domain Clean Evals (Tiers S & A) Legacy Benchmarks (Tiers C & D) Public Lift (Boost)
Agentic & Enterprise60.2p (2)57.1p (1)-3.1p
Coding & SWE59.4p (1)61.2p (1)+1.8p
Math & Scientific Reasoning43.0p (1)58.6p (1)+15.6p
Clean Evals Avg (Tiers S & A)
50.8p
5 private, leak-resistant tests
Legacy Benchmarks Avg (Tiers C & D)
59.0p
3 public or aging tests
Matched Public Lift (Boost)
+4.9p
Near zero or negative indicates clean generalization
Machine-Readable Export

Audited Benchmark Performance of Muse Ember

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

Top-Tier Validated Tests (Tiers S & A) (5)

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

BenchmarkBenchmark Integrity StatusDomainRankPercentile
AutomationBench-AAPrivate / FreshAgentic & Enterprise#17/5067.3p
FrontierCode-AA 2.0Private / FreshCoding & SWE#20/4659.4p
AA-Briefcase v1.1Private / FreshAgentic & Enterprise#24/5053.1p
MLCR-AAPrivate / FreshMath & Scientific Reasoning#28/4643.0p
AA-Omniscience v1.1Private / FreshGeneral Knowledge & QA#29/4740.4p

Mid-Tier Evaluations (Tier B) (1)

Partially exposed or transitional evaluations with moderate leakage risk.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
HLE-DiamondSemi-ExposedGeneral Knowledge & QA#27/4945.8p

Legacy & Vulnerable Benchmarks (Tiers C & D) (3)

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

BenchmarkBenchmark Integrity StatusDomainRankPercentile
Terminal-Bench 4.0Public / AgingCoding & SWE#20/5061.2p
Terminal-Bench-Science 0.1Public / AgingMath & Scientific Reasoning#19/4158.6p
GDP.pdfPublic / AgingAgentic & Enterprise#21/5057.1p