Glossary · Validity and evidence
Task validity
4ValidationThe degree to which tasks elicit the target capability rather than shortcuts, impossible setups, or irrelevant bottlenecks.
Defined in
Handbook · ValidationTask validity and outcome validityDoes success require the capability, and does a pass mean success? The two agent-era links, with checks by output type.Where it comes up
- Evaluation is measurementHandbook · Foundations
- Task distributions and dataset engineeringHandbook · Claim
- The validity argumentHandbook · Validation
- Auditing an evalHandbook · Practice
- CrosswalkHandbook · Reference