{ "artifacts": { "reviewQueueSHA256": "657e3a6696b9ec907e6983b045f240897a3de5035e717ae6a9774d149b98ec3e", "sealedProvenanceSHA256": "25421ba7db52a97b0f72283d0f165670c3701aad74e981c7ac36a5fa5dc3dba7" }, "authoringRequired": 699, "cells": { "en:blessing": 3, "en:complaint": 50, "en:confirmationDecision": 12, "en:followUpReminder": 50, "en:invitation": 50, "en:noWeakIntent": 120, "en:question": 120, "en:replyableOnly": 120, "en:scheduleNegotiation": 50, "en:task": 50, "zh-Hans:complaint": 50, "zh-Hans:followUpReminder": 2, "zh-Hans:question": 12, "zh-Hans:replyableOnly": 4, "zh-Hans:task": 28 }, "clipboardShapeFilter": { "excludedFamilyPrefixes": [ "online_cped_", "online_goemotions_" ], "maximumCharacters": 1500, "minimumCharacters": 25 }, "design": "Single-stage stratified sample. Each record belongs to exactly one (language, stratum) cell; divide by inclusionProbability for prevalence-corrected Horvitz-Thompson metrics.", "humanReviewComplete": false, "languages": { "en": 625, "zh-Hans": 96 }, "policy": "Evaluation-only stratified queue drawn from real licensed text. Weak source labels are stratification input only and are never gold. Never merge these records into training.", "poolRecords": 6939, "records": 721, "rejected": { "duplicateText": 93, "lengthOutsideClipboardShape": 3896, "notRealClipboardShapedSource": 30467 }, "schemaVersion": 1, "seed": 20260902, "sources": [ "ModelTraining/ClipboardSemantics/comprehensive-online-holdout-corpus.jsonl", "ModelTraining/ClipboardSemantics/online-real-holdout-corpus.jsonl" ], "status": "awaiting-double-human-annotation", "unfilledCells": [ { "available": 3, "language": "en", "mustAuthor": 47, "selected": 3, "stratum": "blessing", "target": 50 }, { "available": 12, "language": "en", "mustAuthor": 38, "selected": 12, "stratum": "confirmationDecision", "target": 50 }, { "available": 0, "language": "zh-Hans", "mustAuthor": 50, "selected": 0, "stratum": "blessing", "target": 50 }, { "available": 0, "language": "zh-Hans", "mustAuthor": 50, "selected": 0, "stratum": "confirmationDecision", "target": 50 }, { "available": 0, "language": "zh-Hans", "mustAuthor": 50, "selected": 0, "stratum": "scheduleNegotiation", "target": 50 }, { "available": 0, "language": "zh-Hans", "mustAuthor": 50, "selected": 0, "stratum": "invitation", "target": 50 }, { "available": 2, "language": "zh-Hans", "mustAuthor": 48, "selected": 2, "stratum": "followUpReminder", "target": 50 }, { "available": 28, "language": "zh-Hans", "mustAuthor": 22, "selected": 28, "stratum": "task", "target": 50 }, { "available": 12, "language": "zh-Hans", "mustAuthor": 108, "selected": 12, "stratum": "question", "target": 120 }, { "available": 4, "language": "zh-Hans", "mustAuthor": 116, "selected": 4, "stratum": "replyableOnly", "target": 120 }, { "available": 0, "language": "zh-Hans", "mustAuthor": 120, "selected": 0, "stratum": "noWeakIntent", "target": 120 } ], "validation": { "configuredTrainingOverlap": 0, "containsDetectedPII": false, "duplicateNormalizedTexts": 0 } }