Files
OSGKeyboard/ModelTraining/ClipboardSemantics/random-holdout-report.json
2026-08-27 18:01:46 +08:00

1442 lines
42 KiB
JSON

{
"binaryMacro" : {
"accuracy" : 0.9592,
"f1" : 0.7669,
"precision" : 0.9832,
"recall" : 0.6557
},
"binaryMacroByLanguage" : {
"en" : {
"accuracy" : 0.9507,
"f1" : 0.7085,
"precision" : 0.9787,
"recall" : 0.6085
},
"zh-Hans" : {
"accuracy" : 0.9677,
"f1" : 0.8048,
"precision" : 0.9884,
"recall" : 0.7029
}
},
"classifiers" : [
{
"falseNegativeExamples" : [
{
"confidence" : 0.8754,
"family" : "task_request",
"id" : "random-20260826-en-task_request-004",
"language" : "en",
"text" : "For the record, Design still needs the signature page; please add it before month-end and leave a note in the email chain."
},
{
"confidence" : 0.8676,
"family" : "task_request",
"id" : "random-20260826-en-task_request-014",
"language" : "en",
"text" : "Also, Support still needs the acceptance screenshots; please add it before month-end and leave a note under the ticket."
},
{
"confidence" : 0.8609,
"family" : "task_next_action",
"id" : "random-20260826-en-task_next_action-018",
"language" : "en",
"text" : "One more thing: Your next action is to prepare the retro summary within two business days for Design."
},
{
"confidence" : 0.8124,
"family" : "task_next_action",
"id" : "random-20260826-en-task_next_action-015",
"language" : "en",
"text" : "Your next action is to prepare the test findings before lunch for Design."
},
{
"confidence" : 0.8099,
"family" : "task_request",
"id" : "random-20260826-en-task_request-018",
"language" : "en",
"text" : "Operations still needs the signature page; please add it before month-end and leave a note in the email chain."
}
],
"falsePositiveExamples" : [
],
"id" : "task",
"metrics" : {
"accuracy" : 0.95,
"f1" : 0.835,
"falseNegative" : 34,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.7167,
"total" : 680,
"trueNegative" : 560,
"truePositive" : 86
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9441,
"f1" : 0.8119,
"falseNegative" : 19,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.6833,
"total" : 340,
"trueNegative" : 280,
"truePositive" : 41
},
"zh-Hans" : {
"accuracy" : 0.9559,
"f1" : 0.8571,
"falseNegative" : 15,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.75,
"total" : 340,
"trueNegative" : 280,
"truePositive" : 45
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9691,
"f1" : 0.9136,
"falseNegative" : 9,
"falsePositive" : 12,
"precision" : 0.9024,
"recall" : 0.925,
"total" : 680,
"trueNegative" : 548,
"truePositive" : 111
},
"threshold" : 0.4
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9662,
"f1" : 0.8978,
"falseNegative" : 19,
"falsePositive" : 4,
"precision" : 0.9619,
"recall" : 0.8417,
"total" : 680,
"trueNegative" : 556,
"truePositive" : 101
},
"threshold" : 0.63
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9794,
"f1" : 0.9402,
"falseNegative" : 5,
"falsePositive" : 2,
"precision" : 0.9649,
"recall" : 0.9167,
"total" : 340,
"trueNegative" : 278,
"truePositive" : 55
},
"threshold" : 0.44
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9676,
"f1" : 0.9076,
"falseNegative" : 6,
"falsePositive" : 5,
"precision" : 0.9153,
"recall" : 0.9,
"total" : 340,
"trueNegative" : 275,
"truePositive" : 54
},
"threshold" : 0.47
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9794,
"f1" : 0.9402,
"falseNegative" : 5,
"falsePositive" : 2,
"precision" : 0.9649,
"recall" : 0.9167,
"total" : 340,
"trueNegative" : 278,
"truePositive" : 55
},
"threshold" : 0.44
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9559,
"f1" : 0.8624,
"falseNegative" : 13,
"falsePositive" : 2,
"precision" : 0.9592,
"recall" : 0.7833,
"total" : 340,
"trueNegative" : 278,
"truePositive" : 47
},
"threshold" : 0.66
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.5889,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-005",
"language" : "en",
"text" : "One more thing: The original slot is blocked for me; could we swap to late Friday afternoon?"
},
{
"confidence" : 0.5748,
"family" : "information_question",
"id" : "random-20260826-en-information_question-015",
"language" : "en",
"text" : "Quick check: who owns the final call on access activation?"
},
{
"confidence" : 0.5542,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-006",
"language" : "en",
"text" : "For the record, The original slot is blocked for me; could we swap to after lunch on Thursday?"
},
{
"confidence" : 0.5536,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-002",
"language" : "en",
"text" : "By the way, The original slot is blocked for me; could we swap to after lunch on Thursday?"
},
{
"confidence" : 0.4707,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-007",
"language" : "en",
"text" : "One more thing: The original slot is blocked for me; could we swap to first thing tomorrow?"
}
],
"falsePositiveExamples" : [
],
"id" : "question",
"metrics" : {
"accuracy" : 0.9471,
"f1" : 0.8732,
"falseNegative" : 36,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.775,
"total" : 680,
"trueNegative" : 520,
"truePositive" : 124
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9,
"f1" : 0.7302,
"falseNegative" : 34,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.575,
"total" : 340,
"trueNegative" : 260,
"truePositive" : 46
},
"zh-Hans" : {
"accuracy" : 0.9941,
"f1" : 0.9873,
"falseNegative" : 2,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.975,
"total" : 340,
"trueNegative" : 260,
"truePositive" : 78
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9735,
"f1" : 0.943,
"falseNegative" : 11,
"falsePositive" : 7,
"precision" : 0.9551,
"recall" : 0.9313,
"total" : 680,
"trueNegative" : 513,
"truePositive" : 149
},
"threshold" : 0.06
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9735,
"f1" : 0.943,
"falseNegative" : 11,
"falsePositive" : 7,
"precision" : 0.9551,
"recall" : 0.9313,
"total" : 680,
"trueNegative" : 513,
"truePositive" : 149
},
"threshold" : 0.06
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9588,
"f1" : 0.9079,
"falseNegative" : 11,
"falsePositive" : 3,
"precision" : 0.9583,
"recall" : 0.8625,
"total" : 340,
"trueNegative" : 257,
"truePositive" : 69
},
"threshold" : 0.06
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 1,
"f1" : 1,
"falseNegative" : 0,
"falsePositive" : 0,
"precision" : 1,
"recall" : 1,
"total" : 340,
"trueNegative" : 260,
"truePositive" : 80
},
"threshold" : 0.22
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9588,
"f1" : 0.9079,
"falseNegative" : 11,
"falsePositive" : 3,
"precision" : 0.9583,
"recall" : 0.8625,
"total" : 340,
"trueNegative" : 257,
"truePositive" : 69
},
"threshold" : 0.06
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 1,
"f1" : 1,
"falseNegative" : 0,
"falsePositive" : 0,
"precision" : 1,
"recall" : 1,
"total" : 340,
"trueNegative" : 260,
"truePositive" : 80
},
"threshold" : 0.22
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.7844,
"family" : "fixed_invitation",
"id" : "random-20260826-zh-Hans-fixed_invitation-003",
"language" : "zh-Hans",
"text" : "我给你留了位置,周六傍晚到楼下咖啡店碰面聊方案,能来不?"
},
{
"confidence" : 0.7624,
"family" : "fixed_invitation",
"id" : "random-20260826-zh-Hans-fixed_invitation-001",
"language" : "zh-Hans",
"text" : "简单说,我给你留了位置,今晚九点到园区北门参加庆功会,能来不?"
},
{
"confidence" : 0.7308,
"family" : "fixed_invitation",
"id" : "random-20260826-en-fixed_invitation-005",
"language" : "en",
"text" : "Just to add, I saved you a spot to meet to discuss the proposal Saturday evening at the client office; can you make it?"
},
{
"confidence" : 0.6802,
"family" : "fixed_invitation",
"id" : "random-20260826-zh-Hans-fixed_invitation-017",
"language" : "zh-Hans",
"text" : "刚想到,我给你留了位置,明天十点半到客户办公室碰面聊方案,能来不?"
},
{
"confidence" : 0.6404,
"family" : "fixed_invitation",
"id" : "random-20260826-en-fixed_invitation-007",
"language" : "en",
"text" : "For the record, I saved you a spot to join a small review tonight at nine at the online room; can you make it?"
}
],
"falsePositiveExamples" : [
],
"id" : "invitation",
"metrics" : {
"accuracy" : 0.9735,
"f1" : 0.7097,
"falseNegative" : 18,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.55,
"total" : 680,
"trueNegative" : 640,
"truePositive" : 22
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9676,
"f1" : 0.6207,
"falseNegative" : 11,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.45,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 9
},
"zh-Hans" : {
"accuracy" : 0.9794,
"f1" : 0.7879,
"falseNegative" : 7,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.65,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 13
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9941,
"f1" : 0.95,
"falseNegative" : 2,
"falsePositive" : 2,
"precision" : 0.95,
"recall" : 0.95,
"total" : 680,
"trueNegative" : 638,
"truePositive" : 38
},
"threshold" : 0.37
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9941,
"f1" : 0.95,
"falseNegative" : 2,
"falsePositive" : 2,
"precision" : 0.95,
"recall" : 0.95,
"total" : 680,
"trueNegative" : 638,
"truePositive" : 38
},
"threshold" : 0.37
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 1,
"f1" : 1,
"falseNegative" : 0,
"falsePositive" : 0,
"precision" : 1,
"recall" : 1,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 20
},
"threshold" : 0.12
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9941,
"f1" : 0.95,
"falseNegative" : 1,
"falsePositive" : 1,
"precision" : 0.95,
"recall" : 0.95,
"total" : 340,
"trueNegative" : 319,
"truePositive" : 19
},
"threshold" : 0.4
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 1,
"f1" : 1,
"falseNegative" : 0,
"falsePositive" : 0,
"precision" : 1,
"recall" : 1,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 20
},
"threshold" : 0.12
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9941,
"f1" : 0.95,
"falseNegative" : 1,
"falsePositive" : 1,
"precision" : 0.95,
"recall" : 0.95,
"total" : 340,
"trueNegative" : 319,
"truePositive" : 19
},
"threshold" : 0.4
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.802,
"family" : "complaint_support",
"id" : "random-20260826-zh-Hans-complaint_support-012",
"language" : "zh-Hans",
"text" : "对了,我已经重试三次,还是付款结果重复扣款,导致工作完全卡住。请问什么时候能处理?"
},
{
"confidence" : 0.7754,
"family" : "complaint_support",
"id" : "random-20260826-zh-Hans-complaint_support-015",
"language" : "zh-Hans",
"text" : "单独记一下,附件始终打不开到现在没解决,客户已经在催,能给一个明确答复吗?"
},
{
"confidence" : 0.7686,
"family" : "complaint_support",
"id" : "random-20260826-en-complaint_support-017",
"language" : "en",
"text" : "tracking has not moved for five days is still unresolved and I cannot continue. Can I get a clear answer?"
},
{
"confidence" : 0.7666,
"family" : "complaint_support",
"id" : "random-20260826-zh-Hans-complaint_support-010",
"language" : "zh-Hans",
"text" : "单独记一下,我已经重试三次,还是付款结果重复扣款,导致我无法继续操作。请问什么时候能处理?"
},
{
"confidence" : 0.7311,
"family" : "complaint_support",
"id" : "random-20260826-en-complaint_support-010",
"language" : "en",
"text" : "the account is locked again is still unresolved and delivery is now delayed. Can I get a clear answer?"
}
],
"falsePositiveExamples" : [
],
"id" : "complaint",
"metrics" : {
"accuracy" : 0.9618,
"f1" : 0.5185,
"falseNegative" : 26,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.35,
"total" : 680,
"trueNegative" : 640,
"truePositive" : 14
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9471,
"f1" : 0.1818,
"falseNegative" : 18,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.1,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 2
},
"zh-Hans" : {
"accuracy" : 0.9765,
"f1" : 0.75,
"falseNegative" : 8,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.6,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 12
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9721,
"f1" : 0.7077,
"falseNegative" : 17,
"falsePositive" : 2,
"precision" : 0.92,
"recall" : 0.575,
"total" : 680,
"trueNegative" : 638,
"truePositive" : 23
},
"threshold" : 0.61
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9706,
"f1" : 0.6774,
"falseNegative" : 19,
"falsePositive" : 1,
"precision" : 0.9545,
"recall" : 0.525,
"total" : 680,
"trueNegative" : 639,
"truePositive" : 21
},
"threshold" : 0.67
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9588,
"f1" : 0.4615,
"falseNegative" : 14,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.3,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 6
},
"threshold" : 0.61
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9853,
"f1" : 0.8571,
"falseNegative" : 5,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.75,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 15
},
"threshold" : 0.72
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9588,
"f1" : 0.4615,
"falseNegative" : 14,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.3,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 6
},
"threshold" : 0.61
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9853,
"f1" : 0.8571,
"falseNegative" : 5,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.75,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 15
},
"threshold" : 0.72
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.6518,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-018",
"language" : "en",
"text" : "By the way, The original slot is blocked for me; could we swap to next Monday afternoon?"
},
{
"confidence" : 0.6236,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-011",
"language" : "en",
"text" : "A quick note: The original slot is blocked for me; could we swap to Tuesday at 4:30?"
},
{
"confidence" : 0.5559,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-015",
"language" : "en",
"text" : "I cannot make tomorrow at 10:30. Would first thing tomorrow be workable instead?"
},
{
"confidence" : 0.5363,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-002",
"language" : "en",
"text" : "By the way, The original slot is blocked for me; could we swap to after lunch on Thursday?"
},
{
"confidence" : 0.5026,
"family" : "schedule_negotiation",
"id" : "random-20260826-en-schedule_negotiation-009",
"language" : "en",
"text" : "A quick note: The original slot is blocked for me; could we swap to late Friday afternoon?"
}
],
"falsePositiveExamples" : [
],
"id" : "scheduleNegotiation",
"metrics" : {
"accuracy" : 0.9647,
"f1" : 0.5714,
"falseNegative" : 24,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.4,
"total" : 680,
"trueNegative" : 640,
"truePositive" : 16
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9676,
"f1" : 0.6207,
"falseNegative" : 11,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.45,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 9
},
"zh-Hans" : {
"accuracy" : 0.9618,
"f1" : 0.5185,
"falseNegative" : 13,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.35,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 7
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9603,
"f1" : 0.5091,
"falseNegative" : 26,
"falsePositive" : 1,
"precision" : 0.9333,
"recall" : 0.35,
"total" : 680,
"trueNegative" : 639,
"truePositive" : 14
},
"threshold" : 0.65
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9588,
"f1" : 0.4615,
"falseNegative" : 28,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.3,
"total" : 680,
"trueNegative" : 640,
"truePositive" : 12
},
"threshold" : 0.68
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9676,
"f1" : 0.6452,
"falseNegative" : 10,
"falsePositive" : 1,
"precision" : 0.9091,
"recall" : 0.5,
"total" : 340,
"trueNegative" : 319,
"truePositive" : 10
},
"threshold" : 0.65
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9765,
"f1" : 0.75,
"falseNegative" : 8,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.6,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 12
},
"threshold" : 0.28
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9676,
"f1" : 0.6207,
"falseNegative" : 11,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.45,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 9
},
"threshold" : 0.68
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9765,
"f1" : 0.75,
"falseNegative" : 8,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.6,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 12
},
"threshold" : 0.28
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.7057,
"family" : "confirmation_decision",
"id" : "random-20260826-en-confirmation_decision-005",
"language" : "en",
"text" : "The main point is this: No more comparisons: choose the offline resolution and use it as the final direction."
},
{
"confidence" : 0.6984,
"family" : "confirmation_decision",
"id" : "random-20260826-en-confirmation_decision-004",
"language" : "en",
"text" : "One more thing: I formally approve the lightweight version; start the next steps by tomorrow evening."
},
{
"confidence" : 0.6851,
"family" : "confirmation_decision",
"id" : "random-20260826-zh-Hans-confirmation_decision-015",
"language" : "zh-Hans",
"text" : "对了,不用再比较了,就选季度结算,后续都按这个口径走。"
},
{
"confidence" : 0.6688,
"family" : "confirmation_decision",
"id" : "random-20260826-zh-Hans-confirmation_decision-001",
"language" : "zh-Hans",
"text" : "还有一件事,我正式确认供应商丙,请明晚之前启动后续安排。"
},
{
"confidence" : 0.6408,
"family" : "confirmation_decision",
"id" : "random-20260826-en-confirmation_decision-013",
"language" : "en",
"text" : "For the record, No more comparisons: choose a limited rollout first and use it as the final direction."
}
],
"falsePositiveExamples" : [
{
"confidence" : 0.9098,
"family" : "neutral_fact",
"id" : "random-20260826-en-neutral_fact-015",
"language" : "en",
"text" : "For context, The reference page states that the current version is 2.1."
}
],
"id" : "confirmationDecision",
"metrics" : {
"accuracy" : 0.9779,
"f1" : 0.7761,
"falseNegative" : 14,
"falsePositive" : 1,
"precision" : 0.963,
"recall" : 0.65,
"total" : 680,
"trueNegative" : 639,
"truePositive" : 26
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9794,
"f1" : 0.8,
"falseNegative" : 6,
"falsePositive" : 1,
"precision" : 0.9333,
"recall" : 0.7,
"total" : 340,
"trueNegative" : 319,
"truePositive" : 14
},
"zh-Hans" : {
"accuracy" : 0.9765,
"f1" : 0.75,
"falseNegative" : 8,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.6,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 12
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9882,
"f1" : 0.8974,
"falseNegative" : 5,
"falsePositive" : 3,
"precision" : 0.9211,
"recall" : 0.875,
"total" : 680,
"trueNegative" : 637,
"truePositive" : 35
},
"threshold" : 0.48
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9824,
"f1" : 0.8286,
"falseNegative" : 11,
"falsePositive" : 1,
"precision" : 0.9667,
"recall" : 0.725,
"total" : 680,
"trueNegative" : 639,
"truePositive" : 29
},
"threshold" : 0.68
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9882,
"f1" : 0.9,
"falseNegative" : 2,
"falsePositive" : 2,
"precision" : 0.9,
"recall" : 0.9,
"total" : 340,
"trueNegative" : 318,
"truePositive" : 18
},
"threshold" : 0.57
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9912,
"f1" : 0.9189,
"falseNegative" : 3,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.85,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 17
},
"threshold" : 0.43
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9588,
"f1" : 0.4615,
"falseNegative" : 14,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.3,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 6
},
"threshold" : 0.91
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9912,
"f1" : 0.9189,
"falseNegative" : 3,
"falsePositive" : 0,
"precision" : 1,
"recall" : 0.85,
"total" : 340,
"trueNegative" : 320,
"truePositive" : 17
},
"threshold" : 0.43
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.6584,
"family" : "task_next_action",
"id" : "random-20260826-zh-Hans-task_next_action-001",
"language" : "zh-Hans",
"text" : "顺便提一句,下一步由你整理交付清单,请在午休前交给设计同学。"
},
{
"confidence" : 0.6132,
"family" : "task_next_action",
"id" : "random-20260826-zh-Hans-task_next_action-018",
"language" : "zh-Hans",
"text" : "补充一下,下一步由你整理复盘摘要,请在午休前交给设计同学。"
},
{
"confidence" : 0.5828,
"family" : "task_next_action",
"id" : "random-20260826-zh-Hans-task_next_action-006",
"language" : "zh-Hans",
"text" : "简单说,下一步由你整理预算备注,请在午休前交给设计同学。"
},
{
"confidence" : 0.5773,
"family" : "task_next_action",
"id" : "random-20260826-zh-Hans-task_next_action-005",
"language" : "zh-Hans",
"text" : "下一步由你整理交付清单,请在明晚之前交给财务。"
},
{
"confidence" : 0.5765,
"family" : "task_next_action",
"id" : "random-20260826-zh-Hans-task_next_action-012",
"language" : "zh-Hans",
"text" : "还有一件事,下一步由你整理预算备注,请在明晚之前交给设计同学。"
}
],
"falsePositiveExamples" : [
{
"confidence" : 0.9127,
"family" : "vague_future_boundary",
"id" : "random-20260826-zh-Hans-vague_future_boundary-016",
"language" : "zh-Hans",
"text" : "补充一下,哪天有心情再找医生聊下一步安排吧。"
},
{
"confidence" : 0.8991,
"family" : "vague_future_boundary",
"id" : "random-20260826-zh-Hans-vague_future_boundary-011",
"language" : "zh-Hans",
"text" : "另外,哪天有心情再找法务聊下一步安排吧。"
},
{
"confidence" : 0.8202,
"family" : "information_question",
"id" : "random-20260826-en-information_question-009",
"language" : "en",
"text" : "Quick check: who owns the final call on the release?"
},
{
"confidence" : 0.7834,
"family" : "information_question",
"id" : "random-20260826-en-information_question-008",
"language" : "en",
"text" : "In short, Quick check: who owns the final call on the release?"
}
],
"id" : "followUpReminder",
"metrics" : {
"accuracy" : 0.9706,
"f1" : 0.9123,
"falseNegative" : 16,
"falsePositive" : 4,
"precision" : 0.963,
"recall" : 0.8667,
"total" : 680,
"trueNegative" : 556,
"truePositive" : 104
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9941,
"f1" : 0.9836,
"falseNegative" : 0,
"falsePositive" : 2,
"precision" : 0.9677,
"recall" : 1,
"total" : 340,
"trueNegative" : 278,
"truePositive" : 60
},
"zh-Hans" : {
"accuracy" : 0.9471,
"f1" : 0.8302,
"falseNegative" : 16,
"falsePositive" : 2,
"precision" : 0.9565,
"recall" : 0.7333,
"total" : 340,
"trueNegative" : 278,
"truePositive" : 44
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9706,
"f1" : 0.9174,
"falseNegative" : 9,
"falsePositive" : 11,
"precision" : 0.9098,
"recall" : 0.925,
"total" : 680,
"trueNegative" : 549,
"truePositive" : 111
},
"threshold" : 0.4
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.9706,
"f1" : 0.9123,
"falseNegative" : 16,
"falsePositive" : 4,
"precision" : 0.963,
"recall" : 0.8667,
"total" : 680,
"trueNegative" : 556,
"truePositive" : 104
},
"threshold" : 0.73
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 1,
"f1" : 1,
"falseNegative" : 0,
"falsePositive" : 0,
"precision" : 1,
"recall" : 1,
"total" : 340,
"trueNegative" : 280,
"truePositive" : 60
},
"threshold" : 0.83
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9706,
"f1" : 0.918,
"falseNegative" : 4,
"falsePositive" : 6,
"precision" : 0.9032,
"recall" : 0.9333,
"total" : 340,
"trueNegative" : 274,
"truePositive" : 56
},
"threshold" : 0.29
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 1,
"f1" : 1,
"falseNegative" : 0,
"falsePositive" : 0,
"precision" : 1,
"recall" : 1,
"total" : 340,
"trueNegative" : 280,
"truePositive" : 60
},
"threshold" : 0.83
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9471,
"f1" : 0.8302,
"falseNegative" : 16,
"falsePositive" : 2,
"precision" : 0.9565,
"recall" : 0.7333,
"total" : 340,
"trueNegative" : 278,
"truePositive" : 44
},
"threshold" : 0.73
}
}
},
{
"falseNegativeExamples" : [
{
"confidence" : 0.7472,
"family" : "task_next_action",
"id" : "random-20260826-en-task_next_action-004",
"language" : "en",
"text" : "The main point is this: Your next action is to prepare the delivery checklist before the next review for Operations."
},
{
"confidence" : 0.7332,
"family" : "follow_up_instruction",
"id" : "random-20260826-en-follow_up_instruction-018",
"language" : "en",
"text" : "One more thing: Once the payment lands, check with the landlord the next day about the final decision."
},
{
"confidence" : 0.7327,
"family" : "positive_feedback",
"id" : "random-20260826-zh-Hans-positive_feedback-017",
"language" : "zh-Hans",
"text" : "单独记一下,必须夸一下,处理结果超出预期,谢谢你们。"
},
{
"confidence" : 0.6852,
"family" : "conversation",
"id" : "random-20260826-en-conversation-006",
"language" : "en",
"text" : "For context, Just letting you know, I finally wrapped up today; we can talk later."
},
{
"confidence" : 0.6811,
"family" : "conversation",
"id" : "random-20260826-zh-Hans-conversation-001",
"language" : "zh-Hans",
"text" : "简单说,刚看到你发的照片,突然想到你可能会想知道。"
}
],
"falsePositiveExamples" : [
{
"confidence" : 0.9917,
"family" : "negative_news",
"id" : "random-20260826-en-negative_news-006",
"language" : "en",
"text" : "By the way, Data shows that many flights were delayed overnight, with no request for support."
},
{
"confidence" : 0.9892,
"family" : "follow_up_personal_reminder",
"id" : "random-20260826-zh-Hans-follow_up_personal_reminder-004",
"language" : "zh-Hans",
"text" : "刚想到,个人备忘:提醒我两个工作日内联系客户经理,追一下到账情况。"
},
{
"confidence" : 0.9856,
"family" : "vague_future_boundary",
"id" : "random-20260826-zh-Hans-vague_future_boundary-010",
"language" : "zh-Hans",
"text" : "刚想到,哪天有心情再找客户经理聊交付日期吧。"
},
{
"confidence" : 0.9801,
"family" : "vague_future_boundary",
"id" : "random-20260826-zh-Hans-vague_future_boundary-006",
"language" : "zh-Hans",
"text" : "刚想到,哪天有心情再找医生聊交付日期吧。"
},
{
"confidence" : 0.9663,
"family" : "vague_future_boundary",
"id" : "random-20260826-zh-Hans-vague_future_boundary-013",
"language" : "zh-Hans",
"text" : "对了,哪天有心情再找法务聊下一步安排吧。"
}
],
"id" : "replyableMessage",
"metrics" : {
"accuracy" : 0.9279,
"f1" : 0.9387,
"falseNegative" : 25,
"falsePositive" : 24,
"precision" : 0.9398,
"recall" : 0.9375,
"total" : 680,
"trueNegative" : 256,
"truePositive" : 375
},
"metricsByLanguage" : {
"en" : {
"accuracy" : 0.9059,
"f1" : 0.9192,
"falseNegative" : 18,
"falsePositive" : 14,
"precision" : 0.9286,
"recall" : 0.91,
"total" : 340,
"trueNegative" : 126,
"truePositive" : 182
},
"zh-Hans" : {
"accuracy" : 0.95,
"f1" : 0.9578,
"falseNegative" : 7,
"falsePositive" : 10,
"precision" : 0.9507,
"recall" : 0.965,
"total" : 340,
"trueNegative" : 130,
"truePositive" : 193
}
},
"thresholdAt90Precision" : {
"metrics" : {
"accuracy" : 0.9412,
"f1" : 0.952,
"falseNegative" : 3,
"falsePositive" : 37,
"precision" : 0.9147,
"recall" : 0.9925,
"total" : 680,
"trueNegative" : 243,
"truePositive" : 397
},
"threshold" : 0.46
},
"thresholdAt95Precision" : {
"metrics" : {
"accuracy" : 0.8941,
"f1" : 0.9055,
"falseNegative" : 55,
"falsePositive" : 17,
"precision" : 0.953,
"recall" : 0.8625,
"total" : 680,
"trueNegative" : 263,
"truePositive" : 345
},
"threshold" : 0.87
},
"thresholdsAt90PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.9382,
"f1" : 0.9501,
"falseNegative" : 0,
"falsePositive" : 21,
"precision" : 0.905,
"recall" : 1,
"total" : 340,
"trueNegative" : 119,
"truePositive" : 200
},
"threshold" : 0.46
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.9324,
"f1" : 0.9451,
"falseNegative" : 2,
"falsePositive" : 21,
"precision" : 0.9041,
"recall" : 0.99,
"total" : 340,
"trueNegative" : 119,
"truePositive" : 198
},
"threshold" : 0.32
}
},
"thresholdsAt95PrecisionByLanguage" : {
"en" : {
"metrics" : {
"accuracy" : 0.8441,
"f1" : 0.854,
"falseNegative" : 45,
"falsePositive" : 8,
"precision" : 0.9509,
"recall" : 0.775,
"total" : 340,
"trueNegative" : 132,
"truePositive" : 155
},
"threshold" : 0.87
},
"zh-Hans" : {
"metrics" : {
"accuracy" : 0.95,
"f1" : 0.9578,
"falseNegative" : 7,
"falsePositive" : 10,
"precision" : 0.9507,
"recall" : 0.965,
"total" : 340,
"trueNegative" : 130,
"truePositive" : 193
},
"threshold" : 0.77
}
}
}
],
"corpusRecordCount" : 680,
"exactTrainingOverlapCount" : 0,
"familyCount" : 17,
"generatedAt" : "2026-08-27T01:15:13Z",
"languageCounts" : {
"en" : 340,
"zh-Hans" : 340
},
"manifestSchemaVersion" : 2,
"seed" : 20260826,
"sentiment" : {
"accuracy" : 0.8353,
"correct" : 568,
"macroF1" : 0.6994,
"perLabelF1" : {
"negative" : 0.5841,
"neutral" : 0.9428,
"positive" : 0.5714
},
"total" : 680,
"unknown" : 90,
"unknownRate" : 0.1324
}
}