HOL Throughline
AILuminate Assessment Standard
Evaluation methodology for AI response risk, separate from messaging and protocol interoperability.
Scope & publication status
- Record type
- Specification
- Authority
- community spec
- Native status
- Published
- Normalized publication stage
- Published
- Version examined
- 1.4
- Record basis
- reviewed
Responsibilities: Evaluation & assurance evidence
Parent record: MLCommons
Publication milestones
These are evidence states for individual milestones, not a certification or a percentage of technical completeness.
- Scope & charterevidencedMilestone source
- Normative specificationevidencedMilestone source
- Public reviewunknown
- Conformance testsunknown
- Published releaseevidencedMilestone source
Related work
- HCS-25
liaison candidate · proposed
Explore a common evidence envelope while retaining distinct subjects, methods and trust models.
- Analytics WG
governs · documented
Working group contributes to development of the next assessment standard.
- Safety WG
governs · documented
Working group contributes to development of the next assessment standard.
Questions still open
- Track v1.5 development separately from the published v1.4 assessment standard.
HOL Throughline · Sources before conclusions