Build SCOPE benchmark

README.md

@@ -69,8 +69,8 @@ Per row:
 
 - **`sample`** — the input block, exactly the text the classifier is given. What a block contains differs by task and is described in each section below.
 - **`overall_label`** — the gold answer, always a string.
-- **`rationale`** — why the row carries its label. Generated from the human-annotated samples, not written by the annotators. On `train` only.
-- **auxiliary columns** (optional) — authors' notes, not shown to the classifier. On `test` only.
+- **`rationale`** — why the row carries its label. Generated from the human-annotated samples, not written by the annotators. Filled on `train`; empty on `test`.
+- **`auxiliary`** (some tasks) — authors' notes as a JSON object, not shown to the classifier. Filled on `test`; empty on `train`.
 
 ## Fitness coaching
 
@@ -109,8 +109,9 @@ Inspired by [DynaGuard](https://arxiv.org/abs/2509.02563), this dataset contains
 |---|---|
 | `sample` | The full conversation as a string of alternating <User>...</User> and <Agent>...</Agent> turns; the FINAL turn is the agent's reply being judged. |
 | `overall_label` | The gold answer: `True`, `False`. |
-| `has_user_repetition` (`test` only) | Gold for the first sub-question — did the user re-raise the same unresolved intent at least three times. `1`/`0`. |
-| `is_escalated` (`test` only) | Gold for the second sub-question — did the agent actually escalate. `True`/`False`. |
+| `auxiliary` | Authors' notes as a JSON object string (filled on `test`; empty on `train`), with the keys below. |
+| `auxiliary.has_user_repetition` | Gold for the first sub-question — did the user re-raise the same unresolved intent at least three times. `1`/`0`. |
+| `auxiliary.is_escalated` | Gold for the second sub-question — did the agent actually escalate. `True`/`False`. |
 
 **Rule:** the `message_repetition_with_escalation` entry in [`task_definitions.json`](./task_definitions.json) — `cluster` and `labels`.
 
@@ -136,7 +137,8 @@ The traces are long, with a median length of roughly 29,000 characters and a max
 |---|---|
 | `sample` | The full agent trace as a single JSON-encoded list of {role, content} messages; the FINAL message is the assistant's plan being judged. |
 | `overall_label` | The gold answer: `PASS`, `FAIL`. |
-| `violation_type` (`test` only) | Which failure mode was introduced, for a `FAIL` row; empty for a `PASS` row. |
+| `auxiliary` | Authors' notes as a JSON object string (filled on `test`; empty on `train`), with the keys below. |
+| `auxiliary.violation_type` | Which failure mode was introduced, for a `FAIL` row; empty for a `PASS` row. |
 
 **Rule:** the `plan_verification` entry in [`task_definitions.json`](./task_definitions.json) — `cluster` and `labels`.
 

finance_tile_resolution/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d1879d95e4ec612e3d9d768fa1ea1deadd376774b638035f7591f4d8cf95eee4
-size 50062
+oid sha256:9250d58753c3db7938d91df33fa7f676088bde980bb8d7415983b4e755b9f719
+size 50567

fitness_coaching/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61e57c75e461c4490d6e74faba1b190a4dbe4bc1a18c606d33fbc58b05cbe14d
-size 21749
+oid sha256:8bcef443dcf1db6c0de085e14c43134e81eebf5b5f32cf7147af4c26b17529d9
+size 22252

message_repetition_with_escalation/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4566ebe886d6fe55186163a692465274b1ed46bbd39a1835d459fb2b4175e138
-size 124587
+oid sha256:8e8255ae483bdb276bdaace502577272b81186f99034dbf9e2af60d91b02cf51
+size 124736

message_repetition_with_escalation/train-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:69745de4f0e88babd1e5a870dcbfa276c990367e49be504a7a8ea223b1fd8c6c
-size 83009
+oid sha256:6d07346e4a4d6db69f9ab948014c972bc8f4e2d01ddd57fabeeb384647943dcf
+size 83502

plan_verification/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c1e2ccdb98a3af8e43e5dec69db77dc1af4901f718bc8b795145f2e397010bb
-size 1095127
+oid sha256:958ba2bb404ffd157158fc25887e83b4c36c3325b26df3e65eb50f2639cf54ea
+size 1095790

plan_verification/train-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d647778c46a1b081ebb5656f9696f5bd57f4b01f5b89a06ca74be10e3b8532c
-size 1500295
+oid sha256:18aafbf2d8b2712dcf508811100b3375f6bbb88c13bf874b4f949e5ff88d7bab
+size 1500802