DOC-API-MODEL-schema-1icbr3lltu#overview
ExperimentContext
Dataset-run attribution attached when a scripted evaluation turn is sent. Supported on SendMessageRequest and SendMessageSyncRequest; ordinary conversation traffic leaves it unset.
The attribution must accompany the turn so its observations can be associated with the dataset run; a separate post-run link does not supply this attribution. The dataset is required when experiment context is supplied.
This attribution is separate from client_context, which is included in model prompt context. Do not place evaluation attribution in prompt context.
dataset_item_id is optional: a run can be scored from its traces without assigning each turn to a dataset item. profile_revision_hash supplies a profile revision fingerprint for comparison and attribution.
Meaning and usage
Section ID: DOC-API-MODEL-schema-1icbr3lltu#meaning
See the definition above.
Usage profiles
Section ID: DOC-API-MODEL-schema-1icbr3lltu#fields
| Usage | Fields and validation |
|---|---|
| In requests | View fields |
In requests
Section ID: DOC-API-MODEL-schema-1icbr3lltu#request
API JSON uses JSON field names (camelCase).
| Field | Type | Presence | Meaning |
|---|---|---|---|
Section ID: DOC-API-MODEL-schema-1icbr3lltu#field-datasetidSection ID: DOC-API-MODEL-schema-1icbr3lltu#request-field-datasetiddatasetId | string | Required in requests | Identifier of the dataset to which this run belongs. Required whenever experiment context is supplied. Minimum length: 1. |
Section ID: DOC-API-MODEL-schema-1icbr3lltu#field-runnameSection ID: DOC-API-MODEL-schema-1icbr3lltu#request-field-runnamerunName | string | Required in requests | Names the run. Repeating a run_name for the same dataset appends to that run rather than starting a new one, because the derived experiment id is deterministic — the same property that lets a retried turn land in the run it belongs to. Minimum length: 1. |
Section ID: DOC-API-MODEL-schema-1icbr3lltu#field-datasetitemidSection ID: DOC-API-MODEL-schema-1icbr3lltu#request-field-datasetitemiddatasetItemId | string | Optional | Which dataset item this turn exercises. Optional: a run can be scored on its traces alone, and a harness that does not model items per-turn can leave it empty. |
Section ID: DOC-API-MODEL-schema-1icbr3lltu#field-descriptionSection ID: DOC-API-MODEL-schema-1icbr3lltu#request-field-descriptiondescription | string | Optional | Free-text description displayed with the dataset run. |
Section ID: DOC-API-MODEL-schema-1icbr3lltu#field-metadatajsonSection ID: DOC-API-MODEL-schema-1icbr3lltu#request-field-metadatajsonmetadataJson | string | Optional | Additional run metadata serialized as a JSON object string and carried through unchanged. |
Section ID: DOC-API-MODEL-schema-1icbr3lltu#field-profilerevisionhashSection ID: DOC-API-MODEL-schema-1icbr3lltu#request-field-profilerevisionhashprofileRevisionHash | string | Optional | Profile revision fingerprint attached to the dataset run for comparison and attribution. |
DOC-API-MODEL-schema-1icbr3lltu#validation
Request validation
Section ID: DOC-API-MODEL-schema-1icbr3lltu#request-validation
These rules apply when this value is supplied in a request. Request validation does not guarantee that response fields are present or satisfy the same rules.
| Applies to | Rule | Requirement |
|---|---|---|
datasetId | Minimum length | 1 |
runName | Minimum length | 1 |
Where used
Section ID: DOC-API-MODEL-schema-1icbr3lltu#where-used
| Reference | Relationship |
|---|---|
| Send a message to a conversation | POST /api/v1/llm/send-message |
| Send a message and wait for the result | POST /api/v1/llm/send-message-sync |
Download the public reference contract.
Document ID: DOC-API-MODEL-schema-1icbr3lltu