docs: add base Gemma 4 ablation column to comparison table

README.md

@@ -50,13 +50,13 @@ Evaluated on the horkonpon-corpus held-out eval split (1,037 errorful +
 
 ### Correction
 
-| Metric | GECToR v2 (mt) | GECToR v1 | Gemma 4 E4B |
-|--------|:--------------:|:---------:|:-----------:|
-| **F0.5** | 77.6 | 47.5¹ | **80.8** |
-| Exact match | 51.3% | 18.0%¹ | 65.5% |
-| Precision | 87.6% | 83.8%¹ | 86.3% |
-| Recall | 53.3% | 17.4%¹ | 64.4% |
-| Clean FP | 1.8% | 1.7%¹ | 8.6% |
+| Metric | GECToR v2 (mt) | GECToR v1 | Gemma 4 (FT) | Gemma 4 (base) |
+|--------|:--------------:|:---------:|:-----------:|:--------------:|
+| **F0.5** | 77.6 | 47.5¹ | **80.8** | 2.2 |
+| Exact match | 51.3% | 18.0%¹ | 65.5% | 7.0% |
+| Precision | 87.6% | 83.8%¹ | 86.3% | 1.8% |
+| Recall | 53.3% | 17.4%¹ | 64.4% | 16.5% |
+| Clean FP | 1.8% | 1.7%¹ | 8.6% | 97.2% |
 
 ¹ v1 on horkonpon eval (domain shift — v1 was trained on Elhuyar morphology
 only, so it misses spelling/calque/zalantza errors). v1 scores F0.5=90.2 on its
@@ -67,6 +67,12 @@ own Elhuyar Dem eval set.
 classification. GECToR wins decisively on **clean FP rate** (1.8% vs Gemma 4's
 8.6%) and is lightweight enough for **in-browser ONNX deployment** (~80MB).
 
+The **base Gemma 4 ablation** (F0.5=2.2, 97.2% clean FP) confirms that
+fine-tuning is essential — the base LLM has linguistic knowledge but no
+minimal-edit discipline. See
+[`itzune/gemma-4-e4b-horkonpon`](https://huggingface.co/itzune/gemma-4-e4b-horkonpon)
+for the full ablation report.
+
 ### Error-type classification (new capability)
 
 | Metric | Value |