Mystery_Mail_Guardian / docs /modal-eval-report.json
Auenchanters's picture
docs: Soft Club + eval field notes; eval report committed as evidence
5e547aa
Raw
History Blame Contribute Delete
20.8 kB
{
"n_letters": 48,
"variants": {
"production": {
"prompt_chars": 2634,
"parse_failures": 0,
"level_accuracy": 0.854,
"doc_type_accuracy": 0.417,
"scam_recall": 1.0,
"false_alarm_rate": 0.0,
"amount_hit_rate": 1.0,
"deadline_hit_rate": 1.0,
"mean_latency_s": 7.0,
"mean_output_chars": 639,
"per_case": [
{
"file": "letter_0000.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 6.8
},
{
"file": "letter_0001.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.6
},
{
"file": "letter_0002.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 8.1
},
{
"file": "letter_0003.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 7.4
},
{
"file": "letter_0004.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 7.7
},
{
"file": "letter_0005.png",
"name": "lottery_scam",
"gold": "warning",
"level": "warning",
"doc_type": "charity",
"seconds": 6.2
},
{
"file": "letter_0006.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 7.4
},
{
"file": "letter_0007.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 8.8
},
{
"file": "letter_0008.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.5
},
{
"file": "letter_0009.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.6
},
{
"file": "letter_0010.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.5
},
{
"file": "letter_0011.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 5.1
},
{
"file": "letter_0012.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.4
},
{
"file": "letter_0013.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.6
},
{
"file": "letter_0014.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "marketing",
"seconds": 8.2
},
{
"file": "letter_0015.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 8.3
},
{
"file": "letter_0016.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 9.4
},
{
"file": "letter_0017.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.4
},
{
"file": "letter_0018.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 9.4
},
{
"file": "letter_0019.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 8.2
},
{
"file": "letter_0020.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 8.7
},
{
"file": "letter_0021.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 8.2
},
{
"file": "letter_0022.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 6.1
},
{
"file": "letter_0023.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.8
},
{
"file": "letter_0024.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 8.7
},
{
"file": "letter_0025.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 8.2
},
{
"file": "letter_0026.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 8.0
},
{
"file": "letter_0027.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.6
},
{
"file": "letter_0028.png",
"name": "medical_reminder",
"gold": "low",
"level": "low",
"doc_type": "medical",
"seconds": 6.0
},
{
"file": "letter_0029.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.5
},
{
"file": "letter_0030.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.7
},
{
"file": "letter_0031.png",
"name": "lottery_scam",
"gold": "warning",
"level": "warning",
"doc_type": "charity",
"seconds": 7.4
},
{
"file": "letter_0032.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "government_tax",
"seconds": 7.2
},
{
"file": "letter_0033.png",
"name": "medical_reminder",
"gold": "low",
"level": "low",
"doc_type": "medical",
"seconds": 6.2
},
{
"file": "letter_0034.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 7.8
},
{
"file": "letter_0035.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.5
},
{
"file": "letter_0036.png",
"name": "medical_reminder",
"gold": "low",
"level": "low",
"doc_type": "medical",
"seconds": 5.9
},
{
"file": "letter_0037.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.9
},
{
"file": "letter_0038.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 7.3
},
{
"file": "letter_0039.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.4
},
{
"file": "letter_0040.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.9
},
{
"file": "letter_0041.png",
"name": "lottery_scam",
"gold": "warning",
"level": "warning",
"doc_type": "charity",
"seconds": 7.3
},
{
"file": "letter_0042.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 8.9
},
{
"file": "letter_0043.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 8.6
},
{
"file": "letter_0044.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 6.1
},
{
"file": "letter_0045.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 7.1
},
{
"file": "letter_0046.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 9.4
},
{
"file": "letter_0047.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 8.3
}
]
},
"compact": {
"prompt_chars": 1111,
"parse_failures": 0,
"level_accuracy": 0.708,
"doc_type_accuracy": 0.396,
"scam_recall": 1.0,
"false_alarm_rate": 0.0,
"amount_hit_rate": 1.0,
"deadline_hit_rate": 1.0,
"mean_latency_s": 5.83,
"mean_output_chars": 640,
"per_case": [
{
"file": "letter_0000.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 4.0
},
{
"file": "letter_0001.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.8
},
{
"file": "letter_0002.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 4.1
},
{
"file": "letter_0003.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 6.7
},
{
"file": "letter_0004.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 7.0
},
{
"file": "letter_0005.png",
"name": "lottery_scam",
"gold": "warning",
"level": "warning",
"doc_type": "charity",
"seconds": 5.0
},
{
"file": "letter_0006.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "government_tax",
"seconds": 6.9
},
{
"file": "letter_0007.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.1
},
{
"file": "letter_0008.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.6
},
{
"file": "letter_0009.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.1
},
{
"file": "letter_0010.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.3
},
{
"file": "letter_0011.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 5.4
},
{
"file": "letter_0012.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 4.0
},
{
"file": "letter_0013.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 6.1
},
{
"file": "letter_0014.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.3
},
{
"file": "letter_0015.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "personal",
"seconds": 6.7
},
{
"file": "letter_0016.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 6.3
},
{
"file": "letter_0017.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.8
},
{
"file": "letter_0018.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.0
},
{
"file": "letter_0019.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "government_tax",
"seconds": 6.1
},
{
"file": "letter_0020.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "government_tax",
"seconds": 6.8
},
{
"file": "letter_0021.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.3
},
{
"file": "letter_0022.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 6.5
},
{
"file": "letter_0023.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.5
},
{
"file": "letter_0024.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "warning",
"doc_type": "personal",
"seconds": 6.8
},
{
"file": "letter_0025.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.2
},
{
"file": "letter_0026.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.6
},
{
"file": "letter_0027.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.0
},
{
"file": "letter_0028.png",
"name": "medical_reminder",
"gold": "low",
"level": "low",
"doc_type": "medical",
"seconds": 5.2
},
{
"file": "letter_0029.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 5.7
},
{
"file": "letter_0030.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 3.8
},
{
"file": "letter_0031.png",
"name": "lottery_scam",
"gold": "warning",
"level": "caution",
"doc_type": "charity",
"seconds": 6.8
},
{
"file": "letter_0032.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "government_tax",
"seconds": 7.0
},
{
"file": "letter_0033.png",
"name": "medical_reminder",
"gold": "low",
"level": "low",
"doc_type": "medical",
"seconds": 6.0
},
{
"file": "letter_0034.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.9
},
{
"file": "letter_0035.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 5.7
},
{
"file": "letter_0036.png",
"name": "medical_reminder",
"gold": "low",
"level": "low",
"doc_type": "medical",
"seconds": 5.2
},
{
"file": "letter_0037.png",
"name": "charity_request",
"gold": "low",
"level": "low",
"doc_type": "charity",
"seconds": 4.3
},
{
"file": "letter_0038.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 6.2
},
{
"file": "letter_0039.png",
"name": "phishing_account",
"gold": "warning",
"level": "warning",
"doc_type": "bank",
"seconds": 5.4
},
{
"file": "letter_0040.png",
"name": "bank_notice",
"gold": "low",
"level": "low",
"doc_type": "bank",
"seconds": 5.6
},
{
"file": "letter_0041.png",
"name": "lottery_scam",
"gold": "warning",
"level": "caution",
"doc_type": "charity",
"seconds": 3.5
},
{
"file": "letter_0042.png",
"name": "lookalike_ad",
"gold": "caution",
"level": "caution",
"doc_type": "personal",
"seconds": 6.6
},
{
"file": "letter_0043.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "government_tax",
"seconds": 6.7
},
{
"file": "letter_0044.png",
"name": "phishing_account",
"gold": "warning",
"level": "caution",
"doc_type": "bank",
"seconds": 4.7
},
{
"file": "letter_0045.png",
"name": "utility_bill",
"gold": "low",
"level": "low",
"doc_type": "utility_bill",
"seconds": 7.8
},
{
"file": "letter_0046.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "personal",
"seconds": 6.9
},
{
"file": "letter_0047.png",
"name": "gift_card_scam",
"gold": "warning",
"level": "caution",
"doc_type": "personal",
"seconds": 5.8
}
]
}
}
}