Add results.jsonl so the results table loads in the viewer

README.md

@@ -28,7 +28,7 @@ configs:
 - config_name: results
   data_files:
   - split: test
-    path: results.csv
+    path: results.jsonl
 ---
 
 # Support Routing v1
@@ -46,9 +46,10 @@ The task: given one support message and the fixed policy, select exactly one tea
 | `LABEL-POLICY.md` | Label definitions and precedence rules |
 | `manifest.json` | License, authoring and review procedure, counts and content hash |
 | `results.csv`, `results.json` | One row per model per test case: the recorded answer, status, latency, token counts and, where the provider returns one, the probability for each label |
+| `results.jsonl` | The same result rows, one per line, for the dataset viewer and `datasets` |
 | `results-manifest.json` | The run's conditions, model revisions and file hashes |
 
-Dataset content hash: `sha256:2a2683c23a9cc1aa0d8996762c1b0d53c8bc55d513f8e22a89354098943a0b5b`. Label-policy hash: `sha256:476220aef7fd81e74c78b1453e7558c15feda33bdd965ea4d64e2bf39c18fd42`. These hashes describe the frozen content, not this card or the per-split files.
+Dataset content hash: `sha256:2a2683c23a9cc1aa0d8996762c1b0d53c8bc55d513f8e22a89354098943a0b5b`. Label-policy hash: `sha256:476220aef7fd81e74c78b1453e7558c15feda33bdd965ea4d64e2bf39c18fd42`. These hashes describe the frozen content, not this card or the derived `.jsonl` files.
 
 ## Load it