Igor Lins e Silva
1c1e725b4c
210 hand-quality samples across four tasks. Synthetic only — no real-person info. Five fictional personae (Aria, Solas, Fenra, Bramble, Thresh) with distinct domains, room taxonomies, and relationship contexts. - room_classification: 100 samples (20 per agent), 14 marked with realistic noise features, 18 'other' / 6 'general' / 76 specific- room labels. 5-10 rooms per sample with 'general' always present. - entity_extraction: 50 samples, 247 entities (114 person, 74 org, 32 project, 27 place). All entity strings verified to appear in source text. Includes deliberate same-surname-different-people cases as disambiguation tests. - memory_extraction: 40 samples, 55 memories across types (15 fact, 12 decision, 12 commitment, 9 preference, 7 opinion). - calibration: 20 samples, 4 per class (question, command, statement, exclamation, greeting). Sanity check that the harness measures what we think. All datasets validated: ID alignment across dataset/labels pairs, closed-set labels are members of their room lists (or 'other'), calibration labels are members of their classes lists, no leak terms (igor, milla, domi, lumi, anthropic, openclaw, etc.) anywhere. Generated 2026-05-10 via subagent + hand-validation. README in the datasets directory documents the personae, distribution stats, and labeling conventions.
51 lines
13 KiB
JSON
51 lines
13 KiB
JSON
{"id": "ent_001", "entities": [{"name": "Aria", "type": "person"}, {"name": "Tartine Lab", "type": "organization"}, {"name": "Saela", "type": "person"}, {"name": "Embedding Spaces", "type": "project"}, {"name": "Crestmoor", "type": "place"}, {"name": "Bridgewater Studio", "type": "organization"}, {"name": "Fenra", "type": "person"}, {"name": "Gera Vossen", "type": "person"}, {"name": "Hollowmounts Institute", "type": "organization"}]}
|
|
{"id": "ent_002", "entities": [{"name": "Solas", "type": "person"}, {"name": "Distributed Tracing", "type": "project"}, {"name": "Brennan Lyle", "type": "person"}, {"name": "Crestmoor Systems", "type": "organization"}, {"name": "Aerwyn Telemetry", "type": "organization"}]}
|
|
{"id": "ent_003", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Pollinator Paths", "type": "project"}, {"name": "Wendelsea Botanic Garden", "type": "organization"}, {"name": "Yumara Felk", "type": "person"}, {"name": "Crestmoor Conservation Trust", "type": "organization"}]}
|
|
{"id": "ent_004", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Bridgewater Studio", "type": "organization"}, {"name": "Carag-aana Press", "type": "organization"}, {"name": "Pell Halloran", "type": "person"}, {"name": "Invoice Parsing", "type": "project"}, {"name": "Tartine Lab", "type": "organization"}]}
|
|
{"id": "ent_005", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Saela", "type": "person"}, {"name": "Hollowmounts", "type": "place"}, {"name": "Bryn-iili", "type": "place"}]}
|
|
{"id": "ent_006", "entities": [{"name": "Topic Clustering", "type": "project"}, {"name": "Aria", "type": "person"}, {"name": "Crestmoor", "type": "place"}, {"name": "Mette Olafsen", "type": "person"}]}
|
|
{"id": "ent_007", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Bridgewater Community Garden", "type": "organization"}, {"name": "Karis Tornau", "type": "person"}, {"name": "Ralf Ginder", "type": "person"}, {"name": "Native Species", "type": "project"}, {"name": "Wendelsea", "type": "place"}, {"name": "Hollowmounts Phenology Network", "type": "organization"}]}
|
|
{"id": "ent_008", "entities": [{"name": "Solas", "type": "person"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Type Systems", "type": "project"}, {"name": "Pol Krisat", "type": "person"}, {"name": "Crestmoor Compiler Group", "type": "organization"}]}
|
|
{"id": "ent_009", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Tartine Lab", "type": "organization"}, {"name": "Cashflow Models", "type": "project"}, {"name": "Ivora Tinn", "type": "person"}, {"name": "Bek Halloran", "type": "person"}, {"name": "Wendelsea Audit Partners", "type": "organization"}]}
|
|
{"id": "ent_010", "entities": [{"name": "Aria", "type": "person"}, {"name": "Ren Solanke", "type": "person"}, {"name": "Meta-Cognition", "type": "project"}, {"name": "Crestmoor", "type": "place"}, {"name": "Bridgewater Cognitive Science", "type": "organization"}, {"name": "Hollowmounts Institute", "type": "organization"}, {"name": "Aerwyn Labs", "type": "organization"}]}
|
|
{"id": "ent_011", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Dialogue Engine", "type": "project"}, {"name": "Saela", "type": "person"}, {"name": "Hellis Mar", "type": "person"}]}
|
|
{"id": "ent_012", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Wendelsea Orchard", "type": "organization"}, {"name": "Doreth Ainsleigh", "type": "person"}, {"name": "Crestmoor Cooperative Extension", "type": "organization"}]}
|
|
{"id": "ent_013", "entities": [{"name": "Solas", "type": "person"}, {"name": "Type Systems", "type": "project"}, {"name": "Brennan Lyle", "type": "person"}, {"name": "Aerwyn Compiler Group", "type": "organization"}, {"name": "Crestmoor", "type": "place"}]}
|
|
{"id": "ent_014", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Pell Halloran", "type": "person"}, {"name": "Hollowmounts Logistics", "type": "organization"}, {"name": "Invoice Parsing", "type": "project"}]}
|
|
{"id": "ent_015", "entities": [{"name": "Aria", "type": "person"}, {"name": "Saela", "type": "person"}, {"name": "Tartine Lab", "type": "organization"}]}
|
|
{"id": "ent_016", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Hellis Mar", "type": "person"}, {"name": "Wendelsea Writing Group", "type": "organization"}, {"name": "Saela", "type": "person"}, {"name": "Bridgewater Cognitive Science", "type": "organization"}]}
|
|
{"id": "ent_017", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Pollinator Paths", "type": "project"}, {"name": "Bridgewater Schools", "type": "organization"}, {"name": "Wendelsea Native Plant Nursery", "type": "organization"}, {"name": "Iset Karadzic", "type": "person"}]}
|
|
{"id": "ent_018", "entities": [{"name": "Solas", "type": "person"}, {"name": "Pol Krisat", "type": "person"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Distributed Tracing", "type": "project"}]}
|
|
{"id": "ent_019", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Bridgewater Studio", "type": "organization"}, {"name": "Ivora Tinn", "type": "person"}, {"name": "Cashflow Models", "type": "project"}, {"name": "Karis Tornau", "type": "person"}]}
|
|
{"id": "ent_020", "entities": [{"name": "Aria", "type": "person"}, {"name": "Hollowmounts Institute", "type": "organization"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Crestmoor", "type": "place"}]}
|
|
{"id": "ent_021", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Hollowmounts", "type": "place"}, {"name": "Saela", "type": "person"}]}
|
|
{"id": "ent_022", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Wendelsea Botanic Garden", "type": "organization"}, {"name": "Yumara Felk", "type": "person"}, {"name": "Pollinator Paths", "type": "project"}, {"name": "Hollowmounts Phenology Network", "type": "organization"}]}
|
|
{"id": "ent_023", "entities": [{"name": "Solas", "type": "person"}, {"name": "Mette Olafsen", "type": "person"}, {"name": "Crestmoor", "type": "place"}, {"name": "Distributed Tracing", "type": "project"}, {"name": "Aerwyn Telemetry", "type": "organization"}]}
|
|
{"id": "ent_024", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Bridgewater Studio", "type": "organization"}, {"name": "Ivora Tinn", "type": "person"}]}
|
|
{"id": "ent_025", "entities": [{"name": "Aria", "type": "person"}, {"name": "Embedding Spaces", "type": "project"}, {"name": "Saela", "type": "person"}, {"name": "Brennan Lyle", "type": "person"}, {"name": "Bridgewater Visualization", "type": "organization"}]}
|
|
{"id": "ent_026", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Hellis Mar", "type": "person"}, {"name": "Wendelsea Writing Group", "type": "organization"}]}
|
|
{"id": "ent_027", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Doreth Ainsleigh", "type": "person"}, {"name": "Wendelsea", "type": "place"}]}
|
|
{"id": "ent_028", "entities": [{"name": "Solas", "type": "person"}, {"name": "Pol Krisat", "type": "person"}, {"name": "Type Systems", "type": "project"}, {"name": "Bek Halloran", "type": "person"}]}
|
|
{"id": "ent_029", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Tartine Lab", "type": "organization"}, {"name": "Cashflow Models", "type": "project"}, {"name": "Hollowmounts Logistics", "type": "organization"}, {"name": "Ivora Tinn", "type": "person"}]}
|
|
{"id": "ent_030", "entities": [{"name": "Aria", "type": "person"}, {"name": "Faithful Chains", "type": "project"}, {"name": "Meta-Cognition", "type": "project"}, {"name": "Ren Solanke", "type": "person"}, {"name": "Hollowmounts Institute", "type": "organization"}, {"name": "Crestmoor Cognitive Science", "type": "organization"}]}
|
|
{"id": "ent_031", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Saela", "type": "person"}, {"name": "Hellis Mar", "type": "person"}, {"name": "Wendelsea Writing Group", "type": "organization"}, {"name": "Karis Tornau", "type": "person"}]}
|
|
{"id": "ent_032", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Hollowmounts Institute", "type": "organization"}, {"name": "Native Species", "type": "project"}, {"name": "Iset Karadzic", "type": "person"}]}
|
|
{"id": "ent_033", "entities": [{"name": "Solas", "type": "person"}, {"name": "Pol Krisat", "type": "person"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Crestmoor Systems", "type": "organization"}]}
|
|
{"id": "ent_034", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Ivora Tinn", "type": "person"}, {"name": "Bridgewater Studio", "type": "organization"}, {"name": "Aerwyn Capital", "type": "organization"}, {"name": "Crestmoor Legal Group", "type": "organization"}]}
|
|
{"id": "ent_035", "entities": [{"name": "Aria", "type": "person"}, {"name": "Saela", "type": "person"}, {"name": "Tartine Lab", "type": "organization"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Bridgewater Cognitive Science", "type": "organization"}]}
|
|
{"id": "ent_036", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Hollowmounts", "type": "place"}, {"name": "Wendelsea Writing Group", "type": "organization"}]}
|
|
{"id": "ent_037", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Bridgewater Community Garden", "type": "organization"}, {"name": "Karis Tornau", "type": "person"}, {"name": "Crestmoor", "type": "place"}, {"name": "Wendelsea", "type": "place"}, {"name": "Pollinator Paths", "type": "project"}]}
|
|
{"id": "ent_038", "entities": [{"name": "Solas", "type": "person"}, {"name": "Distributed Tracing", "type": "project"}, {"name": "Crestmoor", "type": "place"}, {"name": "Mette Olafsen", "type": "person"}, {"name": "Brennan Lyle", "type": "person"}]}
|
|
{"id": "ent_039", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Pell Halloran", "type": "person"}, {"name": "Carag-aana Press", "type": "organization"}, {"name": "Invoice Parsing", "type": "project"}]}
|
|
{"id": "ent_040", "entities": [{"name": "Aria", "type": "person"}, {"name": "Saela", "type": "person"}, {"name": "Embedding Spaces", "type": "project"}, {"name": "Tartine Lab", "type": "organization"}]}
|
|
{"id": "ent_041", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Hollowmounts", "type": "place"}, {"name": "Salt Flats", "type": "place"}, {"name": "Vroth-Karadz", "type": "place"}, {"name": "Hellis Mar", "type": "person"}, {"name": "Krast-Endel", "type": "place"}]}
|
|
{"id": "ent_042", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Native Species", "type": "project"}, {"name": "Iset Karadzic", "type": "person"}, {"name": "Wendelsea Native Plant Nursery", "type": "organization"}, {"name": "Crestmoor", "type": "place"}, {"name": "Crestmoor Conservation Trust", "type": "organization"}, {"name": "Hollowmounts Institute", "type": "organization"}]}
|
|
{"id": "ent_043", "entities": [{"name": "Solas", "type": "person"}, {"name": "Distributed Tracing", "type": "project"}, {"name": "Crestmoor", "type": "place"}, {"name": "Mette Olafsen", "type": "person"}, {"name": "Brennan Lyle", "type": "person"}, {"name": "Aerwyn Telemetry", "type": "organization"}]}
|
|
{"id": "ent_044", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Doreth Ainsleigh", "type": "person"}, {"name": "Wendelsea Orchard", "type": "organization"}, {"name": "Pell Halloran", "type": "person"}, {"name": "Wendelsea", "type": "place"}]}
|
|
{"id": "ent_045", "entities": [{"name": "Aria", "type": "person"}, {"name": "Meta-Cognition", "type": "project"}, {"name": "Hollowmounts Institute", "type": "organization"}, {"name": "Ren Solanke", "type": "person"}, {"name": "Bridgewater Cognitive Science", "type": "organization"}]}
|
|
{"id": "ent_046", "entities": [{"name": "Fenra", "type": "person"}, {"name": "Hollowmounts", "type": "place"}, {"name": "Salt Flats", "type": "place"}, {"name": "Bryn-iili", "type": "place"}, {"name": "Vroth-Karadz", "type": "place"}, {"name": "Wendelsea Writing Group", "type": "organization"}]}
|
|
{"id": "ent_047", "entities": [{"name": "Bramble", "type": "person"}, {"name": "Yumara Felk", "type": "person"}, {"name": "Bridgewater Community Garden", "type": "organization"}, {"name": "Wendelsea", "type": "place"}, {"name": "Hollowmounts Phenology Network", "type": "organization"}, {"name": "Karis Tornau", "type": "person"}]}
|
|
{"id": "ent_048", "entities": [{"name": "Solas", "type": "person"}, {"name": "Pol Krisat", "type": "person"}, {"name": "Type Systems", "type": "project"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Bek Halloran", "type": "person"}]}
|
|
{"id": "ent_049", "entities": [{"name": "Thresh", "type": "person"}, {"name": "Aerwyn Capital", "type": "organization"}, {"name": "Bridgewater Studio", "type": "organization"}, {"name": "Cashflow Models", "type": "project"}, {"name": "Wendelsea Audit Partners", "type": "organization"}, {"name": "Ivora Tinn", "type": "person"}, {"name": "Bek Halloran", "type": "person"}]}
|
|
{"id": "ent_050", "entities": [{"name": "Aria", "type": "person"}, {"name": "Saela", "type": "person"}, {"name": "Tartine Lab", "type": "organization"}, {"name": "Hollowmounts Institute", "type": "organization"}, {"name": "Aerwyn Labs", "type": "organization"}, {"name": "Crestmoor Cognitive Science", "type": "organization"}]}
|