File size: 1,720 Bytes
676cb5b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 | ---
library_name: transformers
tags: [model-merging, mergeability, training-free, quotient-merge-distance]
---
# goldfish-en-spa_latn__task_arithmetic__aligned
Training-free merged checkpoint from the **Mergeability** sweep
(`benchmark/emit_lm.py --real`), produced by weight-space merging of two independently
trained parents. No gradient steps were taken.
| field | value |
|---|---|
| pair_id | `goldfish-en-spa_latn` |
| parent_a | `goldfish-models/eng_latn_1000mb` |
| parent_b | `goldfish-models/spa_latn_1000mb` |
| ceiling | `catherinearnett/B-GPT_en_es_simultaneous` |
| operator | `task_arithmetic` |
| alignment | `aligned` |
| align_method | `permutation` |
| regime | `different_corpora` |
| eval_langs | `eng_latn+spa_latn` |
| nll_merge | `6.2756` |
| nll_floor | `4.5501` |
| param_coverage | `1.0` |
| MS | `-2.5372` |
## How it was made
Parents were loaded, activations extracted on a shared calibration corpus, and the merge applied
either **naive** (parents combined in their own coordinates) or **aligned** (parent B carried into
parent A's residual-stream basis via `common.alignment.residual_basis_map` before merging —
permutation for same-width pairs, orthogonal/rectangular for cross-width).
`MS` is the recovery score from `common.eval.mergeability_score` (merged vs. floor vs. ceiling), the
same normalisation used by Zhou et al., so it is comparable across rows of the sweep.
## Caveats
Sub-1B merges are noisy; an aligned signal where the naive one is noise is the finding, not a bug.
Rows without a joint ceiling are floor-relative and must not be read as absolute recovery.
Generated automatically — see the [mergeability repo](https://github.com/suchirsalhan/mergeability).
|