Learn modules
Open reference modules, data requirements, and policy constraints.
Open LearnRESEARCH BENCHMARK
Deterministic evidence for model quality, runtime reliability, and validation gates.
This screen centralizes benchmark outputs, dataset validation gates, globe coverage checks, and load-test proof artifacts in one place.
Learn → Research → Runtime funnel
Use one guided path: training context, benchmark proof, then runtime deployment actions.
Research validation
Review deterministic benchmark deltas, gate passes, and supporting artifacts.
View benchmark outcomesExecute in runtime
Move validated settings into suite runtimes for applied planning workflows.
Launch RuntimeBenchmark task outcomes
Task-level error reduction and latency/cost headroom from reproducible benchmark runs.
| Task | sMAPE reduction | Latency | Latency headroom | Cost headroom |
|---|---|---|---|---|
| Loading benchmark artifacts… | ||||
Task gain profile
Suite coverage pulse
Budget headroom ladder
Validation gates and reproducibility pack
All audit artifacts required to verify the benchmark and runtime map coverage.
Gate status
- Loading
- Checking artifacts…