Tier-1/2 confirmatory analysis — dev tier1_taskvector_e105e3d91386, confirmatory tier1_taskvector_confirmatory_625c54bd186f
| pair | task | eligible (own vector dev) | selected condition (dev acc) | conf: zero / few / own / selected / best control | selected − best control [95% CI] |
|---|---|---|---|---|---|
| qwen3b->qwen1.5b | antonym | yes (0.88) | replace_donor_fewshot | procrustes | L16 (0.72) |
| qwen3b->qwen1.5b | capital | yes (0.86) | ridge | L22 | a1.0 (0.18) |
| qwen3b->qwen1.5b | morph | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| qwen3b->qwen1.5b | addthree | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| qwen3b->qwen1.5b | H6 not supported (0 task(s)) | ||||
| qwen1.5b->qwen3b | antonym | yes (0.88) | procrustes_rescaled | L21 | a1.0 (0.75) |
| qwen1.5b->qwen3b | capital | yes (0.82) | procrustes | L28 | a1.0 (0.73) |
| qwen1.5b->qwen3b | morph | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| qwen1.5b->qwen3b | addthree | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| qwen1.5b->qwen3b | H6 partial (1 task(s)) | ||||
| llama3b->qwen3b | antonym | yes (0.88) | cca | L25 | a1.0 (0.81) |
| llama3b->qwen3b | capital | yes (0.82) | cca | L18 | a1.0 (0.50) |
| llama3b->qwen3b | morph | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| llama3b->qwen3b | addthree | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| llama3b->qwen3b | H6 partial (1 task(s)) | ||||
| qwen3b->llama3b | antonym | yes (0.84) | procrustes | L14 | a1.0 (0.66) |
| qwen3b->llama3b | capital | yes (1.00) | procrustes | L14 | a1.0 (0.05) |
| qwen3b->llama3b | morph | no (0.00) | procrustes | L2 | a1.0 (0.00) |
| qwen3b->llama3b | addthree | yes (0.80) | procrustes | L2 | a1.0 (0.00) |
| qwen3b->llama3b | H6 not supported (0 task(s)) |
Note: CIs are exact paired item bootstraps (2 000 draws) on per-item correctness stored in the confirmatory results.