{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiapuba6sztplhxp3vbahndenhhzue73tsk6a37mjiyldrjzfiumq4",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mlwd3ygw3ov2"
  },
  "path": "/t/openai-evals-bug-unknown-parameter/1379991#post_3",
  "publishedAt": "2026-05-15T21:32:36.000Z",
  "site": "https://community.openai.com",
  "textContent": "I’m seeing this error as well. I’m using Dashboard evals only — not the API/SDK directly.\n\nSetup:\n\n  * New eval created from scratch in the OpenAI Platform Dashboard\n  * CSV upload as test data\n  * “Generate new responses” enabled\n  * Text columns only; no image columns\n  * No “Import as responses” selected\n  * Simple run with a developer/user prompt\n  * Semantic similarity grader\n  * I’ve tried recreating the eval and simplifying the CSV/prompt\n\n\n\nThe error appears when trying to run/review the eval:\n\nUnknown parameter: data_source.sampling_params.text.verbosity\n\nI did not intentionally set a verbosity parameter anywhere in the UI. It looks like the Dashboard may be injecting `data_source.sampling_params.text.verbosity` into the eval run config, but that path is not accepted by the current Evals API surface.",
  "title": "OpenAI Evals Bug Unknown Parameter"
}