docs: show every model per-group score grid uncollapsed

README.md

@@ -2161,8 +2161,9 @@ Score is the macro-average over 25 (group, sequence-length) cells; L50 is the
 largest tested length still reaching 0.5 accuracy for that class; Truncated
 counts completions that exhausted the token budget.
 
-<details>
-<summary><b>Exact scores per model, group and sequence length</b> (accuracy over 20 tasks per cell)</summary>
+### Exact scores per model, group and sequence length
+
+Accuracy over 20 tasks per cell.
 
 **gpt-oss-20b** — score 0.390 (TC0 0.355, NC1 0.413); 258/500 truncated at the 4,096-token budget
 
@@ -2224,8 +2225,6 @@ counts completions that exhausted the token budget.
 | s5 (NC1) | 0.00 | 0.00 | 0.00 | 0.10 | 0.00 |
 | m11 (NC1) | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
 
-</details>
-
 ## Versions
 
 - **v2 (current).** All 94 configs regenerated with per-item deterministic sampling (`gdg export-hf`, master seed 0). Corrects the seven defective v1 configs (`psl2_4`, `psl2_8`, `psl2_9`, `psl3_4`, `q8`, `q16`, `q32`); the other 87 configs keep their v1 element indexing exactly, with freshly sampled rows. Adds `complexity_class` and provenance fields to `metadata.json`, and removes the duplicated `TC0/` and `NC1/` trees.