| # Task Method 20-Result Gap Audit |
|
|
| Generated: `2026-06-20T18:44:28+00:00` |
|
|
| This audit is the explicit gap ledger for the 9-method x 20-task result matrix. |
| It keeps missing cells visible while preserving the rule that a numeric score |
| requires a real task target and source artifact. |
|
|
| ## Score Summary |
|
|
| - Method-task records: `180` |
| - Numeric scored records: `180` |
| - Scoreless records: `0` |
| - Proxy-scored records: `6` |
| - Source matrix: [`docs/data/task_method_20_result_matrix.json`](docs/data/task_method_20_result_matrix.json) |
|
|
| ## Method Coverage |
|
|
| | Method | ID | Scored | Scoreless | Proxy | Status counts | |
| | --- | --- | --- | --- | --- | --- | |
| | Minimal | minimal | 20/20 | 0 | 0 | scored: 20 | |
| | Neural MLP | neural_mlp | 20/20 | 0 | 0 | scored: 20 | |
| | 128ep Aligned Simple | metadata128_simple | 20/20 | 0 | 1 | proxy_scored: 1, scored: 19 | |
| | 128ep Aligned NN | metadata128_neural_mlp | 20/20 | 0 | 1 | proxy_scored: 1, scored: 19 | |
| | 128ep Raw Simple | raw128_simple | 20/20 | 0 | 2 | proxy_scored: 2, scored: 18 | |
| | 128ep Raw NN | raw128_neural_mlp | 20/20 | 0 | 2 | proxy_scored: 2, scored: 18 | |
| | Qwen3-Omni v6 LoRA | qwen3_omni_v6_lora | 20/20 | 0 | 0 | scored: 20 | |
| | Cosmos3-Super Reasoner | cosmos3_super_reasoner | 20/20 | 0 | 0 | scored: 20 | |
| | Cosmos3-Nano Future Window | cosmos3_nano_future_window | 20/20 | 0 | 0 | scored: 20 | |
| |
| ## Gap Classes |
| |
| | Status | Count | Next step | |
| | --- | --- | --- | |
| |
| ## Scoreless Records |
| |
| | Task | Task label | Method | Status | Required evidence | |
| | --- | --- | --- | --- | --- | |
| |
| ## Proxy Records |
| |
| | Task | Task label | Method | Metric | Proxy note | |
| | --- | --- | --- | --- | --- | |
| | 15 | Interaction Text Prediction | 128ep Raw Simple | macro_f1 | documented compact proxy completion for this raw128 task axis | |
| | 15 | Interaction Text Prediction | 128ep Raw NN | macro_f1 | documented compact proxy completion for this raw128 task axis | |
| | 19 | Camera-View Synchronization Retrieval | 128ep Aligned Simple | mrr | paired camera-view embeddings are absent from the 128 JSONL/feature export; metadata features retrieve the synchronized same-window depth/audio block as a documented compact synchronization proxy | |
| | 19 | Camera-View Synchronization Retrieval | 128ep Aligned NN | mrr | paired camera-view embeddings are absent from the 128 JSONL/feature export; metadata features retrieve the synchronized same-window depth/audio block as a documented compact synchronization proxy | |
| | 19 | Camera-View Synchronization Retrieval | 128ep Raw Simple | mrr | documented compact proxy completion for this raw128 task axis | |
| | 19 | Camera-View Synchronization Retrieval | 128ep Raw NN | mrr | documented compact proxy completion for this raw128 task axis | |
| |
| ## Immediate Actions |
| |
| - Keep [`docs/data/task_method_20_gap_audit.json`](docs/data/task_method_20_gap_audit.json) next to the radar and matrix so readers can distinguish scored, proxy-scored, and scoreless cells. |
| - Use [`scripts/omni/score_model_output_probes.py`](scripts/omni/score_model_output_probes.py) to check whether train/validation/test model outputs are present before trying to extend Qwen3/Cosmos to all 20 task contracts. |
| - Use [`scripts/omni/launch_all_task_model_scoring_when_free.sh`](scripts/omni/launch_all_task_model_scoring_when_free.sh) as the guarded waiter for a real all-task scoring command when private GPU capacity is available. |
| |
