pino-source-code / artifacts /three_diagnostics_v8 /diagnostic3_crosswalk_v8_2.json
Matthew Ford
pimt v8 diagnostics and qa artifacts
ac9ca3b
Raw
History Blame Contribute Delete
10.6 kB
{
"classification_counts": {
"crosswalk-gap": 5,
"facet-guard-rejected": 3,
"genuine-ambiguity": 18,
"join-error": 1
},
"crosswalk_additions": {
"almonds": "E",
"citrusy": "C",
"coffee": "E",
"tarry": "T"
},
"crosswalk_additions_count": 4,
"crosswalk_version": "v8.2",
"diagnostic": "D3_masked_despite_text_crosswalk_repair",
"facet_guard_audit_count": 59,
"facet_guard_counts": {
"masked": 14,
"relabelled": 1,
"stands": 44
},
"facet_guard_policy": "Single-token source labels stand only when the matched token is dominant. Secondary/dilution/trace facets are masked unless the source foregrounds a clean alternate class.",
"labels_v8_1_count": 94,
"labels_v8_2_count": 122,
"labels_v8_2_pre_guard_count": 136,
"masked_despite_text_records": 27,
"new_authoritative_character_labels_delta": 28,
"outputs": {
"changelog": "artifacts/odour_character_crosswalk_v8_2_changelog.json",
"crosswalk": "artifacts/odour_character_crosswalk_v8_2.json",
"facet_guard_audit": "data/per_material_signal_inspection_v8_2.jsonl",
"labels": "data/odour_character_labels_v8_2.jsonl"
},
"per_record": [
{
"candidate_additions": [
"almonds"
],
"cas": "100-52-7",
"classification": "crosswalk-gap",
"reason": "v8.2 literal descriptor addition maps source text to one unambiguous class.",
"source_material_name": "BENZALDEHYDE",
"v8_2_classes": [
"E"
]
},
{
"candidate_additions": [],
"cas": "104-67-6",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "gamma-UNDECALACTONE"
},
{
"candidate_additions": [
"citrusy"
],
"cas": "112-44-7",
"classification": "crosswalk-gap",
"reason": "v8.2 literal descriptor addition maps source text to one unambiguous class.",
"source_material_name": "UNDECANAL",
"v8_2_classes": [
"C"
]
},
{
"candidate_additions": [
"tarry"
],
"cas": "120-72-9",
"classification": "facet-guard-rejected",
"dominant_character_assessment": "Matched token 'tarry' is a concentrated-state secondary facet. The work-order adjudication identifies Indole's dominant source character as animalic/faecal, so the material must not inherit T from the incidental tarry facet.",
"final_class_or_masked": "U",
"guard_action": "relabelled",
"matched_token": "tarry",
"matched_token_class": "T",
"reason": "Matched token 'tarry' is a concentrated-state secondary facet. The work-order adjudication identifies Indole's dominant source character as animalic/faecal, so the material must not inherit T from the incidental tarry facet.",
"source_material_name": "INDOLE",
"v8_2_classes": [
"T"
]
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "122-78-1",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "PHENYLACETALDEHYDE"
},
{
"candidate_additions": [],
"cas": "141-25-3",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "RHODINOL"
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "515-03-7",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "SCLAREOL"
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "515-69-5",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "BISABOLOL"
},
{
"candidate_additions": [
"citrusy"
],
"cas": "5392-40-1",
"classification": "crosswalk-gap",
"reason": "v8.2 literal descriptor addition maps source text to one unambiguous class.",
"source_material_name": "CITRAL",
"v8_2_classes": [
"C"
]
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "67-64-1",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "ACETONE"
},
{
"candidate_additions": [],
"cas": "68127-65-1",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "RHODINOL"
},
{
"candidate_additions": [
"coffee"
],
"cas": "75-07-0",
"classification": "facet-guard-rejected",
"dominant_character_assessment": "Matched token 'coffee' appears only in 'in high dilution reminiscent of Coffee or Wine'; the source foregrounds pungent ethereal-nauseating odor, which does not map cleanly to one class.",
"final_class_or_masked": "masked",
"guard_action": "masked",
"matched_token": "coffee",
"matched_token_class": "E",
"reason": "Matched token 'coffee' appears only in 'in high dilution reminiscent of Coffee or Wine'; the source foregrounds pungent ethereal-nauseating odor, which does not map cleanly to one class.",
"source_material_name": "ACETALDEHYDE",
"v8_2_classes": [
"E"
]
},
{
"candidate_additions": [],
"cas": "80-56-8",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "alpha-PINENE"
},
{
"candidate_additions": [],
"cas": "97-54-1",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "iso-EUGENOL"
},
{
"candidate_additions": [],
"cas": "SMILES:C(CN)C(=O)O",
"classification": "join-error",
"reason": "SMILES-only material matched through an over-broad structural alias rather than a material name.",
"source_material_name": "3-iso-BUTYLIDEN"
},
{
"candidate_additions": [],
"cas": "SMILES:CC(=O)OCC(COC(C)=O)OC(C)=O",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "TRIACETIN"
},
{
"candidate_additions": [],
"cas": "SMILES:CC(C)[C@H](N)C(=O)O",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "L-Valine."
},
{
"candidate_additions": [
"citrusy"
],
"cas": "SMILES:CC1=CC=C(C(C)C)CC1",
"classification": "crosswalk-gap",
"reason": "v8.2 literal descriptor addition maps source text to one unambiguous class.",
"source_material_name": "alpha-TERPINENE",
"v8_2_classes": [
"C"
]
},
{
"candidate_additions": [],
"cas": "SMILES:CC1=CCC2CC1C2(C)C",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "alpha-PINENE"
},
{
"candidate_additions": [
"coffee"
],
"cas": "SMILES:CCC=O",
"classification": "facet-guard-rejected",
"dominant_character_assessment": "Matched token 'coffee' appears only in extreme dilution; the source foregrounds a very diffusive, penetrating, suffocating/choking odor, which does not map cleanly to one class.",
"final_class_or_masked": "masked",
"guard_action": "masked",
"matched_token": "coffee",
"matched_token_class": "E",
"reason": "Matched token 'coffee' appears only in extreme dilution; the source foregrounds a very diffusive, penetrating, suffocating/choking odor, which does not map cleanly to one class.",
"source_material_name": "PROPIONALDEHYDE",
"v8_2_classes": [
"E"
]
},
{
"ambiguous_terms_seen": [
"anise",
"anisic"
],
"candidate_additions": [],
"cas": "SMILES:COc1ccccc1",
"classification": "genuine-ambiguity",
"reason": "Source text contains weak, negated, or multi-class descriptor evidence with no dominant single class; masked beats guessed.",
"source_material_name": "ANISOLE"
},
{
"candidate_additions": [],
"cas": "SMILES:Cc1ccc(C(C)C)c(O)c1",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "THYMOL"
},
{
"candidate_additions": [],
"cas": "SMILES:Cc1ccccc1",
"classification": "genuine-ambiguity",
"reason": "Arctander join exists but extracted odour text is empty/OCR-unusable.",
"source_material_name": "TOLUENE"
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "SMILES:O=CCc1ccccc1",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "PHENYLACETALDEHYDE"
},
{
"candidate_additions": [
"tarry"
],
"cas": "SMILES:c1ccc2ccccc2c1",
"classification": "crosswalk-gap",
"reason": "v8.2 literal descriptor addition maps source text to one unambiguous class.",
"source_material_name": "NAPHTHALENE",
"v8_2_classes": [
"T"
]
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "SMILES:c1ccc2ncccc2c1",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "QUINOLINE"
},
{
"ambiguous_terms_seen": [],
"candidate_additions": [],
"cas": "SMILES:c1ccncc1",
"classification": "genuine-ambiguity",
"reason": "No unambiguous v8.2 descriptor token found; broad terms remain masked.",
"source_material_name": "PYRIDINE"
}
],
"pimt_version": "v8"
}