Confirm Action

Are you sure you want to proceed?

Release planning · no model scores

STATE-Bench customer support evidence mapping

Exact V3 collection, facet, metric, lineage, identity and rights mapping for STATE-Bench customer support from Microsoft STATE-Bench.

V3 evidence records

1 collection mapping

A repeated upstream benchmark can map to several collections. Each record below preserves its own task role, facets, capability atoms, subject identity, metrics, lineages and exclusions.

primary evidence · customer-service/statebench/state-bench-customer-support

Customer Support & Service Resolution

Which models are best at customer support and service resolution?

Upstream access
machine-ready
Eligibility
runnable-now
Registry
registered
Max task directness
85%

Collection facets

  • Diagnosis and resolution · hybrid · 25%

    Identify the real issue and give a workable resolution path.

  • Policy boundaries · hybrid · 25%

    Apply policy consistently without fabricating authority or exceptions.

  • De-escalation · hybrid · 25%

    Acknowledge frustration while staying direct and useful.

  • Escalation and handoff · hybrid · 25%

    Escalate with complete context, ownership, and a clear next step.

Capability atoms

workflow_resolutionsupport_ux

Subject identity

Declared:
harnessed-model
Metric partitions:
harnessed-model
Separate partition required:
no
Identity projection:
forbidden

Usable evidence lineages

  • statebench-v0-8-main/harnessed-model

Exact metric contract

MetricLineageSubjectProtocolUse
v0.8.main.customer_support.pass_at_1STATE-Bench Main v0.8 — Customer support — Task completion pass@1statebench-v0-8-mainharnessed-modelstatebench-v0.8.mainusable
v0.8.main.customer_support.mean_ux_scoreSTATE-Bench Main v0.8 — Customer support — Mean UX scorestatebench-v0-8-mainharnessed-modelstatebench-v0.8.mainusable

Exclusions and blockers

No mapping-level exclusion is recorded; the page-level release gate still applies.

Mapping note

Never merge v0.7/v0.8 or main/learning tracks.