| # Card Corpus V2 family roadmap |
|
|
| Rows audited: 229,026 |
| Gate implementation: 52/52 |
| Tokenizer execution: all families |
| Split execution: passed |
|
|
| | Priority | Family | Rows | Behavior | Integrity | Distribution | Near duplicates | Length | Splits | Tokenizer | |
| |---|---|---:|---|---|---|---|---|---|---| |
| | PASS | brainstorming_creativity | 384 | — | — | — | — | — | passed | passed | |
| | PASS | casual_conversation | 52,794 | — | — | — | — | — | passed | passed | |
| | PASS | context_clarification | 9,216 | — | — | — | — | — | passed | passed | |
| | PASS | conversation_empathy | 2,048 | — | — | — | — | — | passed | passed | |
| | PASS | critique_revision | 64 | — | — | — | — | — | passed | passed | |
| | PASS | explanation_learning | 36,040 | — | — | — | — | — | passed | passed | |
| | PASS | extraction_classification | 3,456 | — | — | — | — | — | passed | passed | |
| | PASS | grounded_qa | 4,608 | — | — | — | — | — | passed | passed | |
| | PASS | planning_comparison | 384 | — | — | — | — | — | passed | passed | |
| | PASS | practical_action | 384 | — | — | — | — | — | passed | passed | |
| | PASS | reasoning_verification | 108,000 | — | — | — | — | — | passed | passed | |
| | PASS | safety_uncertainty | 3,072 | — | — | — | — | — | passed | passed | |
| | PASS | summarization_synthesis | 4,096 | — | — | — | — | — | passed | passed | |
| | PASS | troubleshooting | 384 | — | — | — | — | — | passed | passed | |
| | PASS | writing_transformation | 4,096 | — | — | — | — | — | passed | passed | |
|
|
| ## Measured rates |
|
|
| | Family | Prompt copy | Internal repeat | Abstract rubric | Prompt collision | Final collision | |
| |---|---:|---:|---:|---:|---:| |
| | brainstorming_creativity | 0.0% | 0.0% | 0.0% | 0.0% | 9.1% | |
| | casual_conversation | 1.5% | 0.2% | 0.2% | 7.8% | 8.8% | |
| | context_clarification | 0.0% | 0.0% | 0.0% | 6.7% | 5.1% | |
| | conversation_empathy | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | critique_revision | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | explanation_learning | 0.0% | 0.0% | 8.0% | 0.7% | 8.1% | |
| | extraction_classification | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | grounded_qa | 0.0% | 0.0% | 0.0% | 4.5% | 3.5% | |
| | planning_comparison | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | practical_action | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | reasoning_verification | 0.0% | 0.0% | 0.0% | 0.8% | 4.8% | |
| | safety_uncertainty | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | summarization_synthesis | 0.0% | 0.0% | 0.0% | 0.0% | 4.9% | |
| | troubleshooting | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | |
| | writing_transformation | 0.0% | 0.0% | 0.0% | 0.0% | 3.5% | |
|
|
| ## Family actions |
|
|
| ### PASS — brainstorming_creativity |
| |
| |
| ### PASS — casual_conversation |
|
|
|
|
| ### PASS — context_clarification |
| |
| |
| ### PASS — conversation_empathy |
|
|
|
|
| ### PASS — critique_revision |
| |
| |
| ### PASS — explanation_learning |
|
|
|
|
| ### PASS — extraction_classification |
| |
| |
| ### PASS — grounded_qa |
|
|
|
|
| ### PASS — planning_comparison |
| |
| |
| ### PASS — practical_action |
|
|
|
|
| ### PASS — reasoning_verification |
| |
| |
| ### PASS — safety_uncertainty |
|
|
|
|
| ### PASS — summarization_synthesis |
| |
| |
| ### PASS — troubleshooting |
| |
| |
| ### PASS — writing_transformation |
|
|
|
|
|
|
|
|
