Clarify RoboValue training and small-scale evaluation data organization (#5)

- Clarify RoboValue training and small-scale evaluation data organization (ba28e71a6649bf7002e891cb596f2a66eddf4815)
- Copy training files to training-data (1-1000 of 14204) (8b399587cafae34113757bbf53734c1ab5816671)
- Copy training files to training-data (1001-2000 of 14204) (566f37994581b37eed01c2ad5472e191e1ad1582)
- Copy training files to training-data (2001-3000 of 14204) (a60580540d41d25a32af8a00cf6574cc6faeb112)
- Copy training files to training-data (3001-4000 of 14204) (bfb852da4836a791161a9a164e3d21e0932e86d9)
- Copy training files to training-data (4001-5000 of 14204) (ce9945d8251b8e8e6651e2babaa089d5a61d86f0)
- Copy training files to training-data (5001-6000 of 14204) (0c64572725025a17592cf691550ba854304a427e)
- Copy training files to training-data (6001-7000 of 14204) (e32e77e7983f257016754bf4f41e2faccffee922)
- Copy training files to training-data (7001-8000 of 14204) (dffa89c36b35bc676c902fe42c6156c989f76be1)
- Copy training files to training-data (8001-9000 of 14204) (fc77b5d3ec2f13ca7ed10d273521cb3b4c103308)
- Copy training files to training-data (9001-10000 of 14204) (ccf7e71bbfa222a8c612db00d6a7fd85d9d02323)
- Copy training files to training-data (10001-11000 of 14204) (f2d2b9c2b527d26c8bae321b32e975f6df33e143)
- Copy training files to training-data (11001-12000 of 14204) (c9033bf3cede6a1ba61a47892cb85e8bfd864f2f)
- Copy training files to training-data (12001-13000 of 14204) (27de009f6df2fd78bdabaa95423949803f6cce0a)
- Copy training files to training-data (13001-14000 of 14204) (ec789e61efd91e7f678928a7da418a1b5ef6d5f9)
- Copy training files to training-data (14001-14204 of 14204) (f363582a39c05acac792641f1251f2a41c508c35)
- Update training paths, pinned download versions, and paper-alignment documentation (ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e)
- Update training paths, pinned download versions, and paper-alignment documentation (485b48acae209dc4c9a71409f71ffa0028b658df)
- Remove the former training root after verifying all training-data copies (52ca1e42a79f3edde004c0e334bba1ca51eea908)

README.md

@@ -1,5 +1,5 @@
 ---
-pretty_name: RoboValue Training and Evaluation Data
+pretty_name: RoboValue Training and Small-Scale Evaluation Data
 language:
 - en
 - zh
@@ -29,7 +29,7 @@ configs:
 
 <div align="center">
   <img src="https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/resolve/main/assets/robovalue-logo.png" alt="RoboValue" width="420" />
-  <h1>RoboValue Training and Evaluation Data</h1>
+  <h1>RoboValue Training and Small-Scale Evaluation Data</h1>
   <p><strong>Training Demonstrations · Small-Scale Real-World &amp; Simulation Public Test Sets</strong></p>
   <p>
     <a href="https://arxiv.org/abs/2610.26663"><img src="https://img.shields.io/badge/arXiv-Paper-red?logo=arxiv" alt="RoboValue paper on arXiv" /></a>
@@ -71,7 +71,7 @@ Training demonstrations and public test trajectories have separate directories a
 
 | Collection | Location | Scale | Purpose |
 | --- | --- | --- | --- |
-| Simulation / real-world training | `data/<domain>/{full-shot,one-shot}` | Four training subsets listed above | Expert demonstrations for training or reference conditioning |
+| Simulation / real-world training | `training-data/<domain>/{full-shot,one-shot}` | Four training subsets listed above | Expert demonstrations for training or reference conditioning |
 | [Real-world public test set](evaluation/real-world/test-30/README.md) | `evaluation/real-world/test-30` | **4 tasks / 30 trajectories / 27,498 time steps / 90 videos** | Small-scale In-Domain diagnostic evaluation |
 | [Simulation public test set](evaluation/simulation/test-30/README.md) | `evaluation/simulation/test-30` | **4 tasks / 30 trajectories / 17,149 time steps / 90 videos** | Small-scale In-Domain diagnostic evaluation |
 
@@ -88,6 +88,8 @@ The real-world subset contains **12 desk**, **12 flower-and-diffuser**, **3 drum
 
 See the [real-world metric coverage and trajectory index](evaluation/real-world/test-30/README.md#metric-coverage) and [simulation metric coverage and trajectory index](evaluation/simulation/test-30/README.md#metric-coverage). Simulation ED/MS terminal labels and semantic alignment are user-attested; TRR/MEM terminal-label evidence retains the source status of pending independent verification. The simulation README documents these annotation evidence states.
 
+**Task and subtask alignment:** [The paper comparison](TASK_ALIGNMENT.md) maps the eight public-test tasks to the 35 training tasks and documents annotation availability, matching definitions, and wording differences. All 58 supplementary Simulation One-Shot subtask definitions match the provided paper snapshot. Its Press By Number card has a confirmation-order inconsistency; the comparison records it while retaining the released annotations. A [machine-readable crosswalk](task_alignment.json) provides local task indices and paper page references.
+
 
 ## 🟦 2. Simulation Training <a name="simulation"></a>
 
@@ -130,7 +132,7 @@ The simulation dataset covers the following 15 tasks.
 | 13 | Swap Blocks |
 | 14 | Sweep Blocks |
 
-[Full-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/tasks.parquet) · [One-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/tasks.parquet)
+[Full-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/tasks.parquet) · [One-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/tasks.parquet)
 
 </details>
 
@@ -146,7 +148,7 @@ The first seven entries are left-side fields; the last seven are right-side fiel
 | 0–6 | `left_joint_0` … `left_joint_6` |
 | 7–13 | `right_joint_0` … `right_joint_6` |
 
-[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/info.json)
+[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/info.json)
 
 </details>
 
@@ -162,7 +164,7 @@ The first seven entries are left-side fields; the last seven are right-side fiel
 | 0–6 | `left_joint_0` … `left_joint_6` |
 | 7–13 | `right_joint_0` … `right_joint_6` |
 
-[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/info.json)
+[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/info.json)
 
 </details>
 
@@ -177,7 +179,7 @@ Loaded images use **CHW (channels, height, width)** order.
 | `observation.images.cam_left_wrist` | `[3, 480, 640]` |
 | `observation.images.cam_right_wrist` | `[3, 480, 640]` |
 
-[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/info.json)
+[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/info.json)
 
 </details>
 
@@ -189,7 +191,7 @@ Loaded images use **CHW (channels, height, width)** order.
 | ---: | --- | ---: | ---: | --- |
 | **789,612** | **4,500 · AV1 / H.264** | **≈45.47 GB** | ≈8.77 hours | `train: 0:1500` |
 
-[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/simulation/full-shot) · [Download and read](#download-and-reading)
+[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/simulation/full-shot) · [Download and read](#download-and-reading)
 
 ### ◉ 2.3 One-Shot
 
@@ -199,9 +201,9 @@ Loaded images use **CHW (channels, height, width)** order.
 | ---: | --- | ---: | ---: | --- |
 | **8,080** | **45 · AV1** | **≈0.50 GB** | ≈5.39 minutes | `train: 0:15` |
 
-[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/simulation/one-shot) · [Task–episode–row mapping](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/provenance/episode_mapping.json) · [Download and read](#download-and-reading)
+[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/simulation/one-shot) · [Task–episode–row mapping](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/provenance/episode_mapping.json) · [Download and read](#download-and-reading)
 
-**Subtask annotations:** [58 paper-defined segments](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/annotations/subtasks/README.md) label all **8,080 frames**, with one subtask per frame. These annotations accompany the training demonstrations and are separate from benchmark test-set labels. [`dataset_labeler`-compatible JSON](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145/data/simulation/one-shot/annotations/subtasks/dataset_labeler) uses the same subtasks and boundaries as the frame labels.
+**Subtask annotations:** [58 paper-defined segments](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/annotations/subtasks/README.md) label all **8,080 frames**, with one subtask per frame. These annotations accompany the training demonstrations and are separate from benchmark test-set labels. [`dataset_labeler`-compatible JSON](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e/training-data/simulation/one-shot/annotations/subtasks/dataset_labeler) uses the same subtasks and boundaries as the frame labels.
 
 ## 🟧 3. Real-World Training <a name="real-world"></a>
 
@@ -249,7 +251,7 @@ The real-world dataset covers the following 20 tasks.
 | 18 | Arrange Flowers and Reed Diffuser |
 | 19 | Prepare Breakfast |
 
-[Full-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/tasks.parquet) · [One-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/tasks.parquet)
+[Full-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/tasks.parquet) · [One-Shot task metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/tasks.parquet)
 
 </details>
 
@@ -274,7 +276,7 @@ The real-world dataset covers the following 20 tasks.
 
 The table lists key fields; the full 290-dimensional definition is in the metadata below.
 
-[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/info.json)
+[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/info.json)
 
 </details>
 
@@ -299,7 +301,7 @@ The table lists key fields; the full 290-dimensional definition is in the metada
 
 The last two reference orientations are distinct from the end-effector target orientations. Full field definitions are in the metadata below.
 
-[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/info.json)
+[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/info.json)
 
 </details>
 
@@ -314,7 +316,7 @@ Loaded images use **CHW (channels, height, width)** order.
 | `observation.images.cam_left_wrist` | `[3, 528, 640]` |
 | `observation.images.cam_right_wrist` | `[3, 528, 640]` |
 
-[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/info.json)
+[Full-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [One-Shot metadata](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/info.json)
 
 </details>
 
@@ -326,7 +328,7 @@ Loaded images use **CHW (channels, height, width)** order.
 | ---: | --- | ---: | ---: | --- |
 | **1,534,321** | **6,000 · HEVC** | **≈23.13 GB** | ≈14.21 hours | `train: 0:2000` |
 
-[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/full-shot) · [Download and read](#download-and-reading)
+[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/real-world/full-shot) · [Download and read](#download-and-reading)
 
 ### ◉ 3.3 One-Shot
 
@@ -338,7 +340,7 @@ For 19 tasks, the selected trajectory has the **smallest original `episode_id` a
 | ---: | --- | ---: | ---: | --- |
 | **16,362** | **60 · HEVC** | **≈0.22 GB** | ≈9.09 minutes | `train: 0:20` |
 
-[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) · [Task–episode–row mapping](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/provenance/episode_mapping.json) · [Download and read](#download-and-reading)
+[**Browse files →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/real-world/one-shot) · [Task–episode–row mapping](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/provenance/episode_mapping.json) · [Download and read](#download-and-reading)
 
 ## 🗂️ 4. Repository Layout <a name="repository-layout"></a>
 
@@ -346,7 +348,7 @@ For 19 tasks, the selected trajectory has the **smallest original `episode_id` a
 README.md                         English dataset card
 README_zh.md                      Chinese dataset card
 subset_catalog.json               Subset index
-data/
+training-data/
 ├── simulation/
 │   ├── full-shot/                1,500 episodes · 15 tasks
 │   └── one-shot/                 15 episodes · 1 per task
@@ -368,6 +370,8 @@ evaluation/
 
 Each released training or public test subset directory is an independent **LeRobot** dataset root (`root`). Both test subsets have frame references, RGB videos, and English instructions, with evaluation labels alongside them; they have no robot state/action vectors. The primary view is the head camera for real-world and the front camera for simulation.
 
+The repository-level training directory is `training-data/`. Each LeRobot root retains its native `data/` frame-table directory; for example, Simulation One-Shot frame tables are stored under `training-data/simulation/one-shot/data/`.
+
 <details>
 <summary><strong>Training subset files and metadata</strong></summary>
 
@@ -396,10 +400,10 @@ Select one `prefix` to download only that subset. Sizes use decimal GB.
 
 | 🗂️ Subset | ↓ `prefix` | ▦ Episodes / time steps | ▣ Size |
 | --- | --- | ---: | ---: |
-| 🟦 **Simulation Full-Shot** | `data/simulation/full-shot` | 1,500 / 789,612 | ≈45.47 GB |
-| 🟦 **Simulation One-Shot** | `data/simulation/one-shot` | 15 / 8,080 | ≈0.50 GB |
-| 🟧 **Real-World Full-Shot** | `data/real-world/full-shot` | 2,000 / 1,534,321 | ≈23.13 GB |
-| 🟧 **Real-World One-Shot** | `data/real-world/one-shot` | 20 / 16,362 | ≈0.22 GB |
+| 🟦 **Simulation Full-Shot** | `training-data/simulation/full-shot` | 1,500 / 789,612 | ≈45.47 GB |
+| 🟦 **Simulation One-Shot** | `training-data/simulation/one-shot` | 15 / 8,080 | ≈0.50 GB |
+| 🟧 **Real-World Full-Shot** | `training-data/real-world/full-shot` | 2,000 / 1,534,321 | ≈23.13 GB |
+| 🟧 **Real-World One-Shot** | `training-data/real-world/one-shot` | 20 / 16,362 | ≈0.22 GB |
 | **Real-World Public Test** | `evaluation/real-world/test-30` | 30 / 27,498 | ≈0.38 GB |
 | **Simulation Public Test** | `evaluation/simulation/test-30` | 30 / 17,149 | ≈0.595 GB |
 
@@ -421,8 +425,8 @@ from huggingface_hub import snapshot_download
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 
 repo_id = "wzlllong/value-model-benchmark-training-data"
-prefix = "data/simulation/one-shot"
-revision = "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145"
+prefix = "training-data/simulation/one-shot"
+revision = "ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e"
 local_dir = Path(os.environ.get("ROBOVALUE_LOCAL_DIR", "./robovalue-data"))
 
 snapshot = snapshot_download(
@@ -466,11 +470,11 @@ After the Simulation One-Shot example above, join the frame labels on `episode_i
 ```python
 import pyarrow.parquet as pq
 
-annotation_prefix = "data/simulation/one-shot/annotations/subtasks"
+annotation_prefix = "training-data/simulation/one-shot/annotations/subtasks"
 annotation_snapshot = snapshot_download(
     repo_id=repo_id,
     repo_type="dataset",
-    revision="7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
+    revision="ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e",
     allow_patterns=[f"{annotation_prefix}/**"],
     local_dir=local_dir,
 )
@@ -525,7 +529,7 @@ shasum -a 256 -c checksums.sha256
 
 ### 5.5 Download and read a public test subset
 
-Choose `evaluation/real-world/test-30` or `evaluation/simulation/test-30` to download one public test subset. The example is pinned to `578593ef28ee7a2d861ebf38162c928291b0d4a9`, which contains both test domains; downloading and loading use the same revision. The original training-data revision shown above predates this test release.
+Choose `evaluation/real-world/test-30` or `evaluation/simulation/test-30` to download one public test subset. The example is pinned to `578593ef28ee7a2d861ebf38162c928291b0d4a9`, which contains both test domains; downloading and loading use the same revision. Training examples use the separately pinned revision containing the `training-data/` directory.
 
 ```python
 from pathlib import Path

README_zh.md

@@ -1,6 +1,6 @@
 <div align="center">
   <img src="https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/resolve/main/assets/robovalue-logo.png" alt="RoboValue" width="420" />
-  <h1>RoboValue 训练与公开测试数据</h1>
+  <h1>RoboValue 训练与小规模评测数据</h1>
   <p><strong>训练示例 · 小规模真机与仿真公开测试集</strong></p>
   <p>
     <a href="https://arxiv.org/abs/2610.26663"><img src="https://img.shields.io/badge/arXiv-Paper-red?logo=arxiv" alt="RoboValue paper on arXiv" /></a>
@@ -42,7 +42,7 @@ RoboValue 提供面向**机器人价值模型**的专家示例，覆盖 **15 个
 
 | 数据集合 | 目录 | 规模 | 用途 |
 | --- | --- | --- | --- |
-| 仿真／真机训练数据 | `data/<domain>/{full-shot,one-shot}` | 上述四个训练子集 | 用于训练或示例参考输入的专家示范 |
+| 仿真／真机训练数据 | `training-data/<domain>/{full-shot,one-shot}` | 上述四个训练子集 | 用于训练或示例参考输入的专家示范 |
 | [真机公开测试集](evaluation/real-world/test-30/README_zh.md) | `evaluation/real-world/test-30` | **4任务／30轨迹／27,498时间步／90视频** | 小规模 In-Domain 诊断评测 |
 | [仿真公开测试集](evaluation/simulation/test-30/README_zh.md) | `evaluation/simulation/test-30` | **4任务／30轨迹／17,149时间步／90视频** | 小规模 In-Domain 诊断评测 |
 
@@ -59,6 +59,8 @@ RoboValue 提供面向**机器人价值模型**的专家示例，覆盖 **15 个
 
 请查看[真机指标覆盖与逐轨迹索引](evaluation/real-world/test-30/README_zh.md#metric-coverage)和[仿真指标覆盖与逐轨迹索引](evaluation/simulation/test-30/README_zh.md#metric-coverage)。仿真 ED/MS 的终局标签和语义对齐来自用户确认；TRR/MEM 的终局标签保留源标注中“待独立验证”的证据状态，详情见仿真说明。
 
+**任务与子任务一致性：**[论文对照说明](TASK_ALIGNMENT_zh.md)给出8个公开测试任务与35个训练任务的对应关系，列明标注是否存在、定义一致项及措辞差异。仿真 One-Shot 补充标注的58个子任务定义均与提供的PDF版本一致。其中“依数字按按钮”的论文卡片存在确认顺序矛盾，对照说明记录该问题并保留已有标注。[机器可读映射](task_alignment.json)提供局部任务编号和论文页码。
+
 
 ## 🟦 2. 仿真训练数据 <a name="simulation"></a>
 
@@ -101,7 +103,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | 13 | 交换积木位置<br><sub>Swap Blocks</sub> |
 | 14 | 清扫积木<br><sub>Sweep Blocks</sub> |
 
-[Full-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/tasks.parquet) · [One-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/tasks.parquet)
+[Full-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/tasks.parquet) · [One-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/tasks.parquet)
 
 </details>
 
@@ -117,7 +119,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | 0–6 | `left_joint_0` … `left_joint_6` |
 | 7–13 | `right_joint_0` … `right_joint_6` |
 
-[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/info.json)
+[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/info.json)
 
 </details>
 
@@ -133,7 +135,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | 0–6 | `left_joint_0` … `left_joint_6` |
 | 7–13 | `right_joint_0` … `right_joint_6` |
 
-[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/info.json)
+[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/info.json)
 
 </details>
 
@@ -148,7 +150,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的���务示
 | `observation.images.cam_left_wrist` | `[3, 480, 640]` |
 | `observation.images.cam_right_wrist` | `[3, 480, 640]` |
 
-[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/meta/info.json)
+[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/meta/info.json)
 
 </details>
 
@@ -160,7 +162,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | ---: | --- | ---: | ---: | --- |
 | **789,612** | **4,500 个 · AV1 / H.264** | **≈45.47 GB** | ≈8.77 小时 | `train: 0:1500` |
 
-[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/simulation/full-shot) · [下载与读取](#download-and-reading)
+[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/simulation/full-shot) · [下载与读取](#download-and-reading)
 
 ### ◉ 2.3 One-Shot
 
@@ -170,9 +172,9 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | ---: | --- | ---: | ---: | --- |
 | **8,080** | **45 个 · AV1** | **≈0.50 GB** | ≈5.39 分钟 | `train: 0:15` |
 
-[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/simulation/one-shot) · [任务—轨迹—行范围映射](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/provenance/episode_mapping.json) · [下载与读取](#download-and-reading)
+[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/simulation/one-shot) · [任务—轨迹—行范围映射](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/provenance/episode_mapping.json) · [下载与读取](#download-and-reading)
 
-**子任务标注：**[58 个论文子任务区间](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/simulation/one-shot/annotations/subtasks/README.md)覆盖全部 **8,080 帧**，每帧对应一个子任务。标注用于这些训练示例，与评测集标签分开。 另提供 [`dataset_labeler` 兼容 JSON](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145/data/simulation/one-shot/annotations/subtasks/dataset_labeler)，与逐帧标签使用相同的子任务和边界。
+**子任务标注：**[58 个论文子任务区间](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/simulation/one-shot/annotations/subtasks/README.md)覆盖全部 **8,080 帧**，每帧对应一个子任务。标注用于这些训练示例，与评测集标签分开。 另提供 [`dataset_labeler` 兼容 JSON](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e/training-data/simulation/one-shot/annotations/subtasks/dataset_labeler)，与逐帧标签使用相同的子任务和边界。
 
 ## 🟧 3. 真实世界训练数据 <a name="real-world"></a>
 
@@ -220,7 +222,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | 18 | 插花与摆放藤条香薰<br><sub>Arrange Flowers and Reed Diffuser</sub> |
 | 19 | 准备早餐<br><sub>Prepare Breakfast</sub> |
 
-[Full-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/tasks.parquet) · [One-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/tasks.parquet)
+[Full-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/tasks.parquet) · [One-Shot 任务元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/tasks.parquet)
 
 </details>
 
@@ -245,7 +247,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 
 上表列出主要字段；完整 290 维定义见以下元数据。
 
-[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/info.json)
+[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/info.json)
 
 </details>
 
@@ -270,7 +272,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 
 最后两项参考朝向不同于末端目标朝向。完整字段定义见以下元数据。
 
-[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/info.json)
+[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/info.json)
 
 </details>
 
@@ -285,7 +287,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | `observation.images.cam_left_wrist` | `[3, 528, 640]` |
 | `observation.images.cam_right_wrist` | `[3, 528, 640]` |
 
-[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/meta/info.json)
+[Full-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [One-Shot 元数据](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/meta/info.json)
 
 </details>
 
@@ -297,7 +299,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | ---: | --- | ---: | ---: | --- |
 | **1,534,321** | **6,000 个 · HEVC** | **≈23.13 GB** | ≈14.21 小时 | `train: 0:2000` |
 
-[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/full-shot) · [下载与读取](#download-and-reading)
+[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/real-world/full-shot) · [下载与读取](#download-and-reading)
 
 ### ◉ 3.3 One-Shot
 
@@ -309,7 +311,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 | ---: | --- | ---: | ---: | --- |
 | **16,362** | **60 个 · HEVC** | **≈0.22 GB** | ≈9.09 分钟 | `train: 0:20` |
 
-[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/data/real-world/one-shot) · [任务—轨迹—行范围映射](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/one-shot/provenance/episode_mapping.json) · [下载与读取](#download-and-reading)
+[**浏览文件 →**](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/tree/main/training-data/real-world/one-shot) · [任务—轨迹—行范围映射](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/one-shot/provenance/episode_mapping.json) · [下载与读取](#download-and-reading)
 
 ## 🗂️ 4. 仓库目录 <a name="repository-layout"></a>
 
@@ -317,7 +319,7 @@ Full-Shot 与 One-Shot 中，每条 **episode** 都是一条完整的任务示
 README.md                         英文数据集卡片
 README_zh.md                      中文数据集卡片
 subset_catalog.json               子集索引
-data/
+training-data/
 ├── simulation/
 │   ├── full-shot/                1,500 条 · 15 个任务
 │   └── one-shot/                 15 条 · 每任务 1 条
@@ -339,6 +341,8 @@ evaluation/
 
 已发布的训练及公开测试子集目录均为独立的 **LeRobot** 数据根目录（`root`）。两个测试子集保留逐帧引用、RGB视频和英文指令，评测标注另行存放，不含机器人状态／动作向量。真机主视角是头部，仿真主视角是正面。
 
+仓库根目录的训练数据位于 `training-data/`。每个 LeRobot 子集内部保留原生逐帧表目录 `data/`，例如仿真 One-Shot 的帧表位于 `training-data/simulation/one-shot/data/`。
+
 <details>
 <summary><strong>训练子集文件结构与元数据</strong></summary>
 
@@ -367,10 +371,10 @@ evaluation/
 
 | 🗂️ 子集 | ↓ `prefix` | ▦ 轨迹 / 时间步 | ▣ 大小 |
 | --- | --- | ---: | ---: |
-| 🟦 **仿真 Full-Shot** | `data/simulation/full-shot` | 1,500 / 789,612 | ≈45.47 GB |
-| 🟦 **仿真 One-Shot** | `data/simulation/one-shot` | 15 / 8,080 | ≈0.50 GB |
-| 🟧 **真机 Full-Shot** | `data/real-world/full-shot` | 2,000 / 1,534,321 | ≈23.13 GB |
-| 🟧 **真机 One-Shot** | `data/real-world/one-shot` | 20 / 16,362 | ≈0.22 GB |
+| 🟦 **仿真 Full-Shot** | `training-data/simulation/full-shot` | 1,500 / 789,612 | ≈45.47 GB |
+| 🟦 **仿真 One-Shot** | `training-data/simulation/one-shot` | 15 / 8,080 | ≈0.50 GB |
+| 🟧 **真机 Full-Shot** | `training-data/real-world/full-shot` | 2,000 / 1,534,321 | ≈23.13 GB |
+| 🟧 **真机 One-Shot** | `training-data/real-world/one-shot` | 20 / 16,362 | ≈0.22 GB |
 | **真机公开测试集** | `evaluation/real-world/test-30` | 30 / 27,498 | ≈0.38 GB |
 | **仿真公开测试集** | `evaluation/simulation/test-30` | 30 / 17,149 | ≈0.595 GB |
 
@@ -392,8 +396,8 @@ from huggingface_hub import snapshot_download
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 
 repo_id = "wzlllong/value-model-benchmark-training-data"
-prefix = "data/simulation/one-shot"
-revision = "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145"
+prefix = "training-data/simulation/one-shot"
+revision = "ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e"
 local_dir = Path(os.environ.get("ROBOVALUE_LOCAL_DIR", "./robovalue-data"))
 
 snapshot = snapshot_download(
@@ -437,11 +441,11 @@ observation.images.cam_right_wrist (3, 480, 640)
 ```python
 import pyarrow.parquet as pq
 
-annotation_prefix = "data/simulation/one-shot/annotations/subtasks"
+annotation_prefix = "training-data/simulation/one-shot/annotations/subtasks"
 annotation_snapshot = snapshot_download(
     repo_id=repo_id,
     repo_type="dataset",
-    revision="7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
+    revision="ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e",
     allow_patterns=[f"{annotation_prefix}/**"],
     local_dir=local_dir,
 )
@@ -496,7 +500,7 @@ shasum -a 256 -c checksums.sha256
 
 ### 5.5 下载并读取公开测试子集
 
-选择 `evaluation/real-world/test-30` 或 `evaluation/simulation/test-30`，仅下载其中一个公开测试子集。下方示例固定到包含两个测试集的版本 `578593ef28ee7a2d861ebf38162c928291b0d4a9`，下载和加载使用同一提交。上方训练示例的旧版本早于本测试集发布。
+选择 `evaluation/real-world/test-30` 或 `evaluation/simulation/test-30`，仅下载其中一个公开测试子集。下方示例固定到包含两个测试集的版本 `578593ef28ee7a2d861ebf38162c928291b0d4a9`，下载和加载使用同一提交。训练示例另行固定到包含 `training-data/` 新目录的提交。
 
 ```python
 from pathlib import Path

TASK_ALIGNMENT.md

@@ -0,0 +1,191 @@
+# Task and Subtask Alignment with the Paper
+
+[简体中文](TASK_ALIGNMENT_zh.md) · [Machine-readable alignment](task_alignment.json) · [Dataset card](README.md)
+
+The 35 training tasks (15 simulation, 20 real-world) all correspond to the supplied **89-page RoboValue paper**. The eight tasks in the small public test release all have training-task counterparts. Full-Shot and One-Shot use identical task instructions within each domain. The simulation One-Shot supplement has **58/58 English subtask definitions matching the paper**, after removing PDF line wrapping; its 58 segment labels and 8,080 frame labels use those same definitions. Across nonempty public test task tables, **28/30 definitions match the paper exactly**; the two exceptions are the desktop task's `storage box` versus `drawer` wording. This is a semantic and coverage audit; it does not establish that every training/test label set or temporal boundary is identical.
+
+The exact PDF reviewed is *RoboValue: A Fine-Grained Sim-and-Real Benchmark for Unified Evaluation of Robotic Value Models*, linked at [arXiv:2610.26663](https://arxiv.org/abs/2610.26663), SHA-256 `5c9301aaa774a92ccb55832a1b8f1b6deeb93e77c7c450c79419a8c7bae5cfc8`. All page numbers below are 1-based PDF pages and equal the printed page numbers. The audit baseline is repository revision `7c3e8c7f730ac1971b93cdd9efb444a22da81cb5`.
+
+## Annotation availability
+
+Paper §3.4 (p.7) states that the training split does not include subtask boundary annotations and adaptation annotations are separate from test labels. All four native training per-frame schemas have no subtask feature. Only **simulation One-Shot** currently provides a separate `annotations/subtasks/` adaptation supplement; simulation Full-Shot and both real-world training subsets have no public subtask boundaries. Descriptions from the paper can therefore be compared to training task instructions, but absent training boundaries cannot be compared to test boundaries. Different executions also have distinct time axes.
+
+| Training subset | Native subtask feature | Public subtask annotation |
+| --- | --- | --- |
+| [real-world/full-shot](training-data/real-world/full-shot) | No | None |
+| [real-world/one-shot](training-data/real-world/one-shot) | No | None |
+| [simulation/full-shot](training-data/simulation/full-shot) | No | None |
+| [simulation/one-shot](training-data/simulation/one-shot) | No | Supplemental taxonomy + segment/frame labels |
+
+## Shared training/test tasks
+
+Task indices are local to each domain/subset. Training task indices below are the same in that domain's Full-Shot and One-Shot task vocabularies. They are independent of the public test indices.
+
+| Domain / task | Training task | Test task | PDF page | Definition comparison |
+| --- | --- | --- | --- | --- |
+| simulation / Fill Pen Holder | 1 | 0 | 29 | Exact |
+| simulation / Hang Mugs | 3 | 1 | 30 | Exact |
+| simulation / Press By Number | 9 | 2 | 32 | Empty MEM test table |
+| simulation / Swap Blocks | 13 | 3 | 35 | Empty MEM test table |
+| real-world / Organize Desktop Workstation | 7 | 0 | 45 | Two object-wording differences |
+| real-world / Arrange Flowers and Reed Diffuser | 18 | 1 | 41 | Exact |
+| real-world / Strike Drum Five Times | 14 | 2 | 49 | Exact |
+| real-world / Press Button Five Times | 5 | 3 | 47 | Exact |
+
+Seven of the eight test instructions exactly match training; all eight match the PDF. **Press Button Five Times** training says “blue button,” while the test/PDF says “button.” Both require five presses with full release and return. Two additional real-world training wording variants outside this test release are the keypad sequence's expanded “five presses total” wording (p.42) and “Along the way” in the cosmetics task (p.44); requested counts, objects, and order constraints are retained. Exact source strings are available in `task_alignment.json`.
+
+## Detailed instruction and subtask comparison
+
+“Same” below means exact English wording after PDF line-wrap removal. For simulation, the paper column also matches the supplemental training taxonomy. For real-world, it is a **paper reference**, because no public training subtask labels are available. Source test IDs map by explicit task-specific rows below; they are not universal semantic identities across execution orders.
+
+### Fill Pen Holder (simulation, p.29)
+
+**Training/test/paper instruction:** Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Hold the empty pen holder. | 001 | Same |
+| S2 | Insert the first selected pen into the pen holder. | 002 | Same |
+| S3 | Insert the second selected pen into the pen holder. | 003 | Same |
+| S4 | Insert the third selected pen into the pen holder. | 004 | Same |
+| S5 | Insert the remaining pen into the pen holder. | 005 | Same |
+| S6 | Place the filled pen holder upright on the table. | 006 | Same |
+
+### Hang Mugs (simulation, p.30)
+
+**Training/test/paper instruction:** Hang all three mugs on the mug rack.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Hang the first cup on an empty hook. | 001 | Same |
+| S2 | Hang the second cup on another empty hook. | 002 | Same |
+| S3 | Hang the third cup on the remaining empty hook. | 003 | Same |
+
+### Press By Number (simulation, p.32)
+
+**Training/test/paper instruction:** Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Press the first button the required number of times. | — | Empty MEM table; no temporal GT |
+| S2 | Press the blue button. | — | Empty MEM table; no temporal GT |
+| S3 | Press the second button the required number of times. | — | Empty MEM table; no temporal GT |
+| S4 | Press the blue button again. | — | Empty MEM table; no temporal GT |
+
+### Swap Blocks (simulation, p.35)
+
+**Training/test/paper instruction:** Swap the two blocks using the empty mat, pressing the button after each move.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Put the first block on the empty mat. | — | Empty MEM table; no temporal GT |
+| S2 | Press the button to confirm the first move. | — | Empty MEM table; no temporal GT |
+| S3 | Transfer the second block onto the original first block’s position. | — | Empty MEM table; no temporal GT |
+| S4 | Press the button to confirm the second move. | — | Empty MEM table; no temporal GT |
+| S5 | Put the first block on the original second block’s mat. | — | Empty MEM table; no temporal GT |
+| S6 | Press the button to confirm the last move. | — | Empty MEM table; no temporal GT |
+
+### Organize Desktop Workstation (real-world, p.45)
+
+**Training/test/paper instruction:** Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Place the pen in the pen holder. | 001 | Same |
+| S2 | Place the eraser in the drawer. | 002 | Place the eraser in the storage box. |
+| S3 | Close the drawer. | 003 | Close the storage box. |
+| S4 | Turn off the desk lamp. | 004 | Same |
+| S5 | Unplug the desk lamp. | 005 | Same |
+
+### Arrange Flowers and Reed Diffuser (real-world, p.41)
+
+**Training/test/paper instruction:** Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Insert the first flower into the vase. | 001 | Same |
+| S2 | Insert the second flower into the vase. | 002 | Same |
+| S3 | Insert the third flower into the vase. | 003 | Same |
+| S4 | Place the reed diffuser bottle on the coaster. | 004 | Same |
+| S5 | Insert the diffuser reed into the reed diffuser bottle. | 005 | Same |
+
+### Strike Drum Five Times (real-world, p.49)
+
+**Training/test/paper instruction:** Pick up the drumstick and strike the center of the drumhead 5 times continuously.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Strike the center of the drum for the first time. | 001 | Same |
+| S2 | Strike the center of the drum for the second time. | 002 | Same |
+| S3 | Strike the center of the drum for the third time. | 003 | Same |
+| S4 | Strike the center of the drum for the fourth time. | 004 | Same |
+| S5 | Strike the center of the drum for the fifth time. | 005 | Same |
+| S6 | Place the drumstick on the table. | 006 | Same |
+
+### Press Button Five Times (real-world, p.47)
+
+**Training:** Press the blue button 5 times, fully releasing it and returning to the waiting position after each press.
+
+**Test / paper:** Press the button 5 times, fully releasing it and returning to the waiting position after each press.
+
+| Paper ID | Paper definition (simulation also has training supplement) | Test ID | Public test definition |
+| --- | --- | --- | --- |
+| S1 | Press the button for the first time. | 001 | Same |
+| S2 | Press the button for the second time. | 002 | Same |
+| S3 | Press the button for the third time. | 003 | Same |
+| S4 | Press the button for the fourth time. | 004 | Same |
+| S5 | Press the button for the fifth time. | 005 | Same |
+
+## Order, semantic identity, and unresolved protocol wording
+
+- **Fill Pen Holder (p.29):** hold the holder before four insertions and place it upright after all insertions; pen order may vary. **Hang Mugs (p.30):** mug order may vary. The simulation test preserves ordinal stage labels and a separate object/action/arm semantic overlay for CSVC. The same local order slot need not identify the same pen or mug across solutions.
+- **Flowers and diffuser (p.41):** the three flowers are consecutive in any flower order; the two diffuser actions may occur in either order around the flower block but cannot interrupt it. Per-episode flower identity remains in the original annotation.
+- **Desktop (p.45):** placing the eraser precedes closing its drawer; switching off the lamp precedes unplugging it. The original test uses “storage box” in two subtask definitions while its instruction and the paper use “drawer.” This comparison preserves both wordings and all original endpoints.
+- **Repeated press/strike (pp.47,49):** preserve full history and the existing sampling anchors; task definitions do not create new framewise counting ground truth. Simulation MEM task and interval tables remain empty.
+- **Press By Number (p.32) requires author protocol clarification:** the instruction, procedure, and order paragraph say to complete both red-button counts before blue confirmation, while printed S1-S4 list first-red → blue → second-red → blue again. This is an internal inconsistency in the card's confirmation count/order. The training taxonomy faithfully reproduces the printed S list; the test instruction follows the printed instruction. Neither original labels nor empty MEM intervals are rewritten to resolve it.
+
+Paper §3.4 (p.7) allows execution-order units to group multiple annotated subtasks **without changing the evaluation definitions or endpoints**. CSVC (§4.3.4, p.12) aligns semantic identities and compares boundary-local gains; grouping, arm roles, and object identity must remain explicit. SIA (§4.3.1, p.10) queries the active stage within annotated intervals. Table 7 (p.26) assigns SIA/CSVC to ED/MS and Memory-VOC to MEM. These public test subsets provide partial diagnostic coverage: simulation retains ten formal metric entries and the original FPL fields, while real-world retains eleven.
+
+## Complete training-task coverage
+
+The following 35 names match the paper task cards. “Exact” compares training task-instruction text; “Wording variant” identifies the three real-world variants described above. This table does not assert the availability of training subtask boundaries.
+
+| Domain | Training task | Paper task title | PDF page | Paper definitions | Instruction |
+| --- | --- | --- | --- | --- | --- |
+| simulation | 0 | Arrange Largest Number | 29 | 4 | Exact |
+| simulation | 1 | Fill Pen Holder | 29 | 6 | Exact |
+| simulation | 2 | Fold Clothes | 30 | 3 | Exact |
+| simulation | 3 | Hang Mugs | 30 | 3 | Exact |
+| simulation | 4 | Insert Tubes | 31 | 3 | Exact |
+| simulation | 5 | Make Toast | 31 | 3 | Exact |
+| simulation | 6 | Organize Table | 32 | 4 | Exact |
+| simulation | 7 | Pack Objects Into Box | 33 | 5 | Exact |
+| simulation | 8 | Play Stacking Toy | 34 | 4 | Exact |
+| simulation | 9 | Press By Number | 32 | 4 | Exact |
+| simulation | 10 | Put Bottles Into Dustbin | 33 | 4 | Exact |
+| simulation | 11 | Stack Blocks | 34 | 3 | Exact |
+| simulation | 12 | Store Laptop And Headphones | 35 | 3 | Exact |
+| simulation | 13 | Swap Blocks | 35 | 6 | Exact |
+| simulation | 14 | Sweep Blocks | 36 | 3 | Exact |
+| real-world | 0 | Transfer Blocks Using Plate | 49 | 4 | Exact |
+| real-world | 1 | Enter Keypad Number Sequence | 42 | 5 | Wording variant |
+| real-world | 2 | Set the Table | 48 | 4 | Exact |
+| real-world | 3 | Arrange Seasoning Jars | 41 | 3 | Exact |
+| real-world | 4 | Load Plastic Plates Into Rack | 43 | 3 | Exact |
+| real-world | 5 | Press Button Five Times | 47 | 5 | Wording variant |
+| real-world | 6 | Place Pot On Induction Cooker | 46 | 4 | Exact |
+| real-world | 7 | Organize Desktop Workstation | 45 | 5 | Exact |
+| real-world | 8 | Organize Books And Pen | 44 | 5 | Exact |
+| real-world | 9 | Arrange Family Tea Set | 40 | 5 | Exact |
+| real-world | 10 | Operate Drawers In Sequence | 43 | 6 | Exact |
+| real-world | 11 | Sort Blocks in Order | 48 | 3 | Exact |
+| real-world | 12 | Fill And Place Pen Holder | 42 | 5 | Exact |
+| real-world | 13 | Organize Cosmetics Storage Box | 44 | 7 | Wording variant |
+| real-world | 14 | Strike Drum Five Times | 49 | 6 | Exact |
+| real-world | 15 | Apply Dish Soap Repeatedly | 40 | 7 | Exact |
+| real-world | 16 | Season Meat Skewers Repeatedly | 47 | 6 | Exact |
+| real-world | 17 | Pack Items For Shipping | 45 | 3 | Exact |
+| real-world | 18 | Arrange Flowers and Reed Diffuser | 41 | 5 | Exact |
+| real-world | 19 | Prepare Breakfast | 46 | 6 | Exact |
+
+The alignment files preserve the distinction between source annotation and paper reference. Native data, test labels, video frames, timestamps, interval endpoints, object assignments, and left/right arm roles remain unchanged.

TASK_ALIGNMENT_zh.md

@@ -0,0 +1,191 @@
+# 训练、公开测试与论文的任务和子任务对照
+
+[English](TASK_ALIGNMENT.md) · [机器可读对照](task_alignment.json) · [仓库中文首页](README_zh.md)
+
+训练集35个任务（仿真15、真机20）均能对应所审查的 **89页RoboValue论文**，本次小规模公开测试集的8个任务均有明确训练任务对应。同域Full-Shot与One-Shot的任务指令完全一致。仿真One-Shot补充标注的 **58/58个英文子任务定义与论文逐字一致**（消除PDF换行），58个区间和8,080行逐帧标签使用相同词条。公开测试非空任务表共有30个定义，**28/30项与论文逐字一致**；两处差异是书桌任务的 `storage box` 与 `drawer` 对象用词。该对照核查语义及覆盖范围，不能据此声称所有训练／测试标签或时间边界完全相同。
+
+所审查PDF标题为 *RoboValue: A Fine-Grained Sim-and-Real Benchmark for Unified Evaluation of Robotic Value Models*，论文入口为 [arXiv:2610.26663](https://arxiv.org/abs/2610.26663)，SHA-256为 `5c9301aaa774a92ccb55832a1b8f1b6deeb93e77c7c450c79419a8c7bae5cfc8`。以下页码均为从1起的PDF页码，与页内印刷页码一致。数据审查基准为仓库版本 `7c3e8c7f730ac1971b93cdd9efb444a22da81cb5`。
+
+## 标注可用范围
+
+论文§3.4（第7页）明确训练划分不含子任务边界，模型适配时构建的标注与测试标签分开。四套原生训练逐帧schema均没有subtask特征。目前只有 **仿真One-Shot** 提供独立 `annotations/subtasks/` 适配补标；仿真Full-Shot和两套真机训练均没有公开子任务边界。因此可将论文定义与训练任务指令进行语义比较，不能比较不存在的训练边界；不同轨迹本来也有各自的时间轴。
+
+| 训练子集 | 原生subtask特征 | 公开子任务标注 |
+| --- | --- | --- |
+| [real-world/full-shot](training-data/real-world/full-shot) | 无 | 无 |
+| [real-world/one-shot](training-data/real-world/one-shot) | 无 | 无 |
+| [simulation/full-shot](training-data/simulation/full-shot) | 无 | 无 |
+| [simulation/one-shot](training-data/simulation/one-shot) | 无 | 补充taxonomy及区间／逐帧标签 |
+
+## 训练与测试共有的8个任务
+
+编号各自局部使用。同域Full-Shot与One-Shot的训练task编号一致，公开测试编号独立于训练编号。
+
+| 域／任务 | 训练task | 测试task | PDF页 | 定义对照 |
+| --- | --- | --- | --- | --- |
+| simulation / Fill Pen Holder | 1 | 0 | 29 | 逐字一致 |
+| simulation / Hang Mugs | 3 | 1 | 30 | 逐字一致 |
+| simulation / Press By Number | 9 | 2 | 32 | MEM测试表为空 |
+| simulation / Swap Blocks | 13 | 3 | 35 | MEM测试表为空 |
+| real-world / Organize Desktop Workstation | 7 | 0 | 45 | 两处对象措辞差异 |
+| real-world / Arrange Flowers and Reed Diffuser | 18 | 1 | 41 | 逐字一致 |
+| real-world / Strike Drum Five Times | 14 | 2 | 49 | 逐字一致 |
+| real-world / Press Button Five Times | 5 | 3 | 47 | 逐字一致 |
+
+8条测试指令中，7条与训练逐字一致，8条均与PDF一致。**Press Button Five Times** 训练指令多了“blue”颜色限定，测试和论文使用“button”；两者都要求按5次、每次完全释放并回到等待位置。另有两项未包含在test-30中的真机训练措辞差别：第42页键盘序列在训练中展开为“五次总计”，第44页化妆品指令在训练中增加“Along the way”；次数、对象及顺序约束保留。完整原文见 `task_alignment.json`。
+
+## 英文指令和逐子任务对照
+
+下表“相同”指消除PDF换行后英文逐字一致。仿真论文列同时对应训练补标taxonomy；真机论文列是 **论文参考定义**，因为训练暂无公开子任务标签。下表逐任务明确编号映射，局部序号不能跨执行顺序直接视为统一物体语义。
+
+### Fill Pen Holder (simulation, p.29)
+
+**训练／测试／论文指令:** Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Hold the empty pen holder. | 001 | 相同 |
+| S2 | Insert the first selected pen into the pen holder. | 002 | 相同 |
+| S3 | Insert the second selected pen into the pen holder. | 003 | 相同 |
+| S4 | Insert the third selected pen into the pen holder. | 004 | 相同 |
+| S5 | Insert the remaining pen into the pen holder. | 005 | 相同 |
+| S6 | Place the filled pen holder upright on the table. | 006 | 相同 |
+
+### Hang Mugs (simulation, p.30)
+
+**训练／测试／论文指令:** Hang all three mugs on the mug rack.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Hang the first cup on an empty hook. | 001 | 相同 |
+| S2 | Hang the second cup on another empty hook. | 002 | 相同 |
+| S3 | Hang the third cup on the remaining empty hook. | 003 | 相同 |
+
+### Press By Number (simulation, p.32)
+
+**训练／测试／论文指令:** Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Press the first button the required number of times. | — | MEM表为空，无该段时间真值 |
+| S2 | Press the blue button. | — | MEM表为空，无该段时间真值 |
+| S3 | Press the second button the required number of times. | — | MEM表为空，无该段时间真值 |
+| S4 | Press the blue button again. | — | MEM表为空，无该段时间真值 |
+
+### Swap Blocks (simulation, p.35)
+
+**训练／测试／论文指令:** Swap the two blocks using the empty mat, pressing the button after each move.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Put the first block on the empty mat. | — | MEM表为空，无该段时间真值 |
+| S2 | Press the button to confirm the first move. | — | MEM表为空，无该段时间真值 |
+| S3 | Transfer the second block onto the original first block’s position. | — | MEM表为空，无该段时间真值 |
+| S4 | Press the button to confirm the second move. | — | MEM表为空，无该段时间真值 |
+| S5 | Put the first block on the original second block’s mat. | — | MEM表为空，无该段时间真值 |
+| S6 | Press the button to confirm the last move. | — | MEM表为空，无该段时间真值 |
+
+### Organize Desktop Workstation (real-world, p.45)
+
+**训练／测试／论文指令:** Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Place the pen in the pen holder. | 001 | 相同 |
+| S2 | Place the eraser in the drawer. | 002 | Place the eraser in the storage box. |
+| S3 | Close the drawer. | 003 | Close the storage box. |
+| S4 | Turn off the desk lamp. | 004 | 相同 |
+| S5 | Unplug the desk lamp. | 005 | 相同 |
+
+### Arrange Flowers and Reed Diffuser (real-world, p.41)
+
+**训练／测试／论文指令:** Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Insert the first flower into the vase. | 001 | 相同 |
+| S2 | Insert the second flower into the vase. | 002 | 相同 |
+| S3 | Insert the third flower into the vase. | 003 | 相同 |
+| S4 | Place the reed diffuser bottle on the coaster. | 004 | 相同 |
+| S5 | Insert the diffuser reed into the reed diffuser bottle. | 005 | 相同 |
+
+### Strike Drum Five Times (real-world, p.49)
+
+**训练／测试／论文指令:** Pick up the drumstick and strike the center of the drumhead 5 times continuously.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Strike the center of the drum for the first time. | 001 | 相同 |
+| S2 | Strike the center of the drum for the second time. | 002 | 相同 |
+| S3 | Strike the center of the drum for the third time. | 003 | 相同 |
+| S4 | Strike the center of the drum for the fourth time. | 004 | 相同 |
+| S5 | Strike the center of the drum for the fifth time. | 005 | 相同 |
+| S6 | Place the drumstick on the table. | 006 | 相同 |
+
+### Press Button Five Times (real-world, p.47)
+
+**训练：** Press the blue button 5 times, fully releasing it and returning to the waiting position after each press.
+
+**测试／论文：** Press the button 5 times, fully releasing it and returning to the waiting position after each press.
+
+| 论文ID | 论文定义（仿真另有训练补标） | 测试ID | 公开测试定义 |
+| --- | --- | --- | --- |
+| S1 | Press the button for the first time. | 001 | 相同 |
+| S2 | Press the button for the second time. | 002 | 相同 |
+| S3 | Press the button for the third time. | 003 | 相同 |
+| S4 | Press the button for the fourth time. | 004 | 相同 |
+| S5 | Press the button for the fifth time. | 005 | 相同 |
+
+## 顺序、语义身份与尚待校准的协议表述
+
+- **笔筒（第29页）：** 先拿住笔筒，完成四次插笔后直立放下，选笔顺序可变。**马克杯（第30页）：** 挂杯顺序可变。仿真测试保留原顺序槽位标签，同时单独保留供CSVC使用的物体／动作／左右臂语义覆盖层；不同成功方案中，同一局部序号未必对应同一支笔或同一个杯子。
+- **花艺与香薰（第41页）：** 三次插花须连续，花色顺序可变；两个香薰动作顺序可变，可在插花块前后执行，不能插入三次插花之间。各轨迹花色归属继续保存在原标注中。
+- **书桌（第45页）：** 放入橡皮后关抽屉，关灯后拔插头。测试两个子任务词条使用“storage box”，任务指令和论文使用“drawer”；对照保留两套原词和原区间端点。
+- **重复按压／敲击（第47、49页）：** 保留完整历史及已有采样锚点；任务描述不等于新增逐帧计数真值。仿真MEM的子任务表和区间表继续为空。
+- **按数字按钮（第32页）需作者校准协议：** 指令、流程及顺序段要求先完成两个红按钮的次数，再按蓝按钮确认；印刷S1-S4表却列为“第一个红按钮→蓝确认→第二个红按钮→蓝确认再次”。这是任务卡中确认次数／顺序的内部不一致。训练taxonomy忠实保留印刷S表，测试指令忠实保留印刷指令；本对照不会通过重写原标签或填充空MEM区间来消除矛盾。
+
+论文§3.4（第7页）允许执行排序单元合并多个标注子任务，但 **评测子任务定义及时间端点保持独立**。CSVC（§4.3.4，第12页）按语义身份比较边界附近价值增益，需要明确排序单元、左右臂角色和物体身份。SIA（§4.3.1，第10页）在标注区间内识别当前阶段。第26页表7规定SIA／CSVC使用ED／MS，Memory-VOC使用MEM。本次测试只是部分诊断覆盖：仿真保留10个正式指标入口及源FPL字段，真机保留11个。
+
+## 全部训练任务覆盖
+
+以下35个任务名均对应论文任务卡。“一致”比较训练任务指令文字，“措辞差别”对应上文三处真机训练变体；本表不表示训练中一定有子任务时间边界。
+
+| 域 | 训练task | 论文任务名 | PDF页 | 论文定义数 | 指令 |
+| --- | --- | --- | --- | --- | --- |
+| simulation | 0 | Arrange Largest Number | 29 | 4 | 一致 |
+| simulation | 1 | Fill Pen Holder | 29 | 6 | 一致 |
+| simulation | 2 | Fold Clothes | 30 | 3 | 一致 |
+| simulation | 3 | Hang Mugs | 30 | 3 | 一致 |
+| simulation | 4 | Insert Tubes | 31 | 3 | 一致 |
+| simulation | 5 | Make Toast | 31 | 3 | 一致 |
+| simulation | 6 | Organize Table | 32 | 4 | 一致 |
+| simulation | 7 | Pack Objects Into Box | 33 | 5 | 一致 |
+| simulation | 8 | Play Stacking Toy | 34 | 4 | 一致 |
+| simulation | 9 | Press By Number | 32 | 4 | 一致 |
+| simulation | 10 | Put Bottles Into Dustbin | 33 | 4 | 一致 |
+| simulation | 11 | Stack Blocks | 34 | 3 | 一致 |
+| simulation | 12 | Store Laptop And Headphones | 35 | 3 | 一致 |
+| simulation | 13 | Swap Blocks | 35 | 6 | 一致 |
+| simulation | 14 | Sweep Blocks | 36 | 3 | 一致 |
+| real-world | 0 | Transfer Blocks Using Plate | 49 | 4 | 一致 |
+| real-world | 1 | Enter Keypad Number Sequence | 42 | 5 | 措辞差别 |
+| real-world | 2 | Set the Table | 48 | 4 | 一致 |
+| real-world | 3 | Arrange Seasoning Jars | 41 | 3 | 一致 |
+| real-world | 4 | Load Plastic Plates Into Rack | 43 | 3 | 一致 |
+| real-world | 5 | Press Button Five Times | 47 | 5 | 措辞差别 |
+| real-world | 6 | Place Pot On Induction Cooker | 46 | 4 | 一致 |
+| real-world | 7 | Organize Desktop Workstation | 45 | 5 | 一致 |
+| real-world | 8 | Organize Books And Pen | 44 | 5 | 一致 |
+| real-world | 9 | Arrange Family Tea Set | 40 | 5 | 一致 |
+| real-world | 10 | Operate Drawers In Sequence | 43 | 6 | 一致 |
+| real-world | 11 | Sort Blocks in Order | 48 | 3 | 一致 |
+| real-world | 12 | Fill And Place Pen Holder | 42 | 5 | 一致 |
+| real-world | 13 | Organize Cosmetics Storage Box | 44 | 7 | 措辞差别 |
+| real-world | 14 | Strike Drum Five Times | 49 | 6 | 一致 |
+| real-world | 15 | Apply Dish Soap Repeatedly | 40 | 7 | 一致 |
+| real-world | 16 | Season Meat Skewers Repeatedly | 47 | 6 | 一致 |
+| real-world | 17 | Pack Items For Shipping | 45 | 3 | 一致 |
+| real-world | 18 | Arrange Flowers and Reed Diffuser | 41 | 5 | 一致 |
+| real-world | 19 | Prepare Breakfast | 46 | 6 | 一致 |
+
+对照文件区分源标注和论文参考。原生数据、测试标签、视频帧、时间戳、区间端点、物体归属和左右臂角色保持原记录。

evaluation/simulation/README.md

@@ -2,7 +2,7 @@
 
 [简体中文](README_zh.md) · [Repository overview](../../README.md)
 
-Public simulation diagnostic subsets are stored here, separately from training demonstrations under `data/simulation/`.
+Public simulation diagnostic subsets are stored here, separately from training demonstrations under `training-data/simulation/`.
 
 | Subset | Tasks | Trajectories | Time steps | Videos | Format |
 | --- | ---: | ---: | ---: | ---: | --- |

evaluation/simulation/README_zh.md

@@ -2,7 +2,7 @@
 
 [English](README.md) · [仓库总览](../../README_zh.md)
 
-本目录存放仿真公开诊断测试子集，与 `data/simulation/` 下的训练示范分开。
+本目录存放仿真公开诊断测试子集，与 `training-data/simulation/` 下的训练示范分开。
 
 | 子集 | 任务 | 轨迹 | 时间步 | 视频 | 格式 |
 | --- | ---: | ---: | ---: | ---: | --- |

subset_catalog.json

@@ -39,7 +39,7 @@
       "tasks": 15,
       "episodes_per_task": 100,
       "episodes": 1500,
-      "path_in_repo": "data/simulation/full-shot",
+      "path_in_repo": "training-data/simulation/full-shot",
       "name_en": "Simulation Full-Shot Training Data",
       "name_zh": "仿真 Full-Shot 训练集",
       "format": "LeRobot v3.0",
@@ -115,7 +115,10 @@
         "收纳笔记本电脑与耳机",
         "交换积木位置",
         "清扫积木"
-      ]
+      ],
+      "role": "training",
+      "size_bytes": 45468403617,
+      "size_scope": "entire subset directory, including documentation and integrity files"
     },
     {
       "subset_id": "simulation-one-shot-data",
@@ -125,7 +128,7 @@
       "tasks": 15,
       "episodes_per_task": 1,
       "episodes": 15,
-      "path_in_repo": "data/simulation/one-shot",
+      "path_in_repo": "training-data/simulation/one-shot",
       "name_en": "Simulation One-Shot Training Data",
       "name_zh": "仿真 One-Shot 训练集",
       "format": "LeRobot v3.0",
@@ -141,7 +144,7 @@
       "video_codecs": [
         "av1"
       ],
-      "episode_mapping": "data/simulation/one-shot/provenance/episode_mapping.json",
+      "episode_mapping": "training-data/simulation/one-shot/provenance/episode_mapping.json",
       "videos": 45,
       "platform": "ARX",
       "source_description_en": "Based on open-source RoboDojo data reviewed by our team, with selected demonstrations recorded again.",
@@ -199,8 +202,8 @@
       ],
       "annotations": {
         "subtasks": {
-          "path_in_repo": "data/simulation/one-shot/annotations/subtasks",
-          "revision": "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
+          "path_in_repo": "training-data/simulation/one-shot/annotations/subtasks",
+          "revision": "ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e",
           "status": "user_confirmed_release",
           "target": "active_semantic_subtask",
           "scope": "training_demonstrations",
@@ -216,35 +219,35 @@
             "frame_index"
           ],
           "global_index_matches": "index",
-          "frame_labels": "data/simulation/one-shot/annotations/subtasks/frame_labels.parquet",
-          "taxonomy": "data/simulation/one-shot/annotations/subtasks/taxonomy.json",
+          "frame_labels": "training-data/simulation/one-shot/annotations/subtasks/frame_labels.parquet",
+          "taxonomy": "training-data/simulation/one-shot/annotations/subtasks/taxonomy.json",
           "size_bytes": 146350,
           "payload_size_bytes": 146350,
-          "documentation": "data/simulation/one-shot/annotations/subtasks/README.md",
+          "documentation": "training-data/simulation/one-shot/annotations/subtasks/README.md",
           "interval_convention": "[start_frame,end_frame)",
           "boundary_mask": "boundary_uncertain",
           "payload_files": 24,
           "dataset_labeler": {
-            "path_in_repo": "data/simulation/one-shot/annotations/subtasks/dataset_labeler",
+            "path_in_repo": "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler",
             "format": "dataset_labeler_sia_cspc",
             "task_files": [
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/arrange_largest_number_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/fill_pen_holder_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/fold_clothes_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/hang_mugs_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/insert_tubes_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/make_toast_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/organize_table_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/pack_objects_into_box_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/play_stacking_toy_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/press_by_number_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/put_bottles_into_dustbin_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/stack_blocks_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/store_laptop_and_headphones_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/swap_blocks_sia_cspc_annotations.json",
-              "data/simulation/one-shot/annotations/subtasks/dataset_labeler/sweep_blocks_sia_cspc_annotations.json"
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/arrange_largest_number_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/fill_pen_holder_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/fold_clothes_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/hang_mugs_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/insert_tubes_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/make_toast_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/organize_table_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/pack_objects_into_box_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/play_stacking_toy_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/press_by_number_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/put_bottles_into_dustbin_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/stack_blocks_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/store_laptop_and_headphones_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/swap_blocks_sia_cspc_annotations.json",
+              "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/sweep_blocks_sia_cspc_annotations.json"
             ],
-            "exceptional_intervals": "data/simulation/one-shot/annotations/subtasks/dataset_labeler/exceptional_intervals.json",
+            "exceptional_intervals": "training-data/simulation/one-shot/annotations/subtasks/dataset_labeler/exceptional_intervals.json",
             "transfer_export_records": 15,
             "transfer_export_boundary_points": 73,
             "record_fields": [
@@ -271,9 +274,14 @@
             "metric": "SIA+CSPC",
             "group": "one-shot",
             "gap": 20
-          }
+          },
+          "source_revision": "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
+          "directory_size_bytes": 158894
         }
-      }
+      },
+      "role": "training",
+      "size_bytes": 499726709,
+      "size_scope": "entire subset directory, including documentation and integrity files"
     },
     {
       "subset_id": "real-world-full-shot-data",
@@ -282,7 +290,7 @@
       "name_en": "Real-World Full-Shot Training Data",
       "name_zh": "真机 Full-Shot 训练集",
       "availability": "available",
-      "path_in_repo": "data/real-world/full-shot",
+      "path_in_repo": "training-data/real-world/full-shot",
       "format": "LeRobot v3.0",
       "robot_type": "G2A",
       "tasks": 20,
@@ -347,8 +355,10 @@
         "插花与摆放藤条香薰",
         "准备早餐"
       ],
-      "size_bytes": 23125415964,
-      "duration_seconds": 51144.03333333333
+      "size_bytes": 23126335697,
+      "duration_seconds": 51144.03333333333,
+      "role": "training",
+      "size_scope": "entire subset directory, including documentation and integrity files"
     },
     {
       "subset_id": "real-world-one-shot-data",
@@ -358,7 +368,7 @@
       "tasks": 20,
       "episodes_per_task": 1,
       "episodes": 20,
-      "path_in_repo": "data/real-world/one-shot",
+      "path_in_repo": "training-data/real-world/one-shot",
       "name_en": "Real-World One-Shot Training Data",
       "name_zh": "真机 One-Shot 训练集",
       "format": "LeRobot v3.0",
@@ -374,11 +384,11 @@
       "video_codecs": [
         "hevc"
       ],
-      "episode_mapping": "data/real-world/one-shot/provenance/episode_mapping.json",
+      "episode_mapping": "training-data/real-world/one-shot/provenance/episode_mapping.json",
       "source_episode_selection": "minimum original episode_id among the 100 source episodes for 19 tasks; tea uses the minimum original episode_id from the 27-episode tea refresh batch",
       "videos": 60,
       "platform": "AgiBot Genie02",
-      "size_bytes": 215642018,
+      "size_bytes": 215666142,
       "duration_seconds": 545.4,
       "display_task_names": [
         "Transfer Blocks Using Plate",
@@ -434,7 +444,9 @@
           "source_episode_id": 52375079,
           "selection": "minimum original episode_id in the 27-episode tea refresh batch"
         }
-      ]
+      ],
+      "role": "training",
+      "size_scope": "entire subset directory, including documentation and integrity files"
     },
     {
       "subset_id": "real-world-test-30",
@@ -518,7 +530,8 @@
       ],
       "primary_camera": "observation.images.top_head",
       "primary_video_column": "video_primary",
-      "data_revision": "578593ef28ee7a2d861ebf38162c928291b0d4a9"
+      "data_revision": "578593ef28ee7a2d861ebf38162c928291b0d4a9",
+      "size_scope": "entire subset directory, including documentation and integrity files"
     },
     {
       "subset_id": "simulation-test-30",
@@ -602,14 +615,28 @@
       "primary_video_column": "video_primary",
       "annotation_evidence_note": "ED/MS terminal labels and semantic alignment are user-attested; TRR/MEM terminal-label evidence remains pending independent verification.",
       "size_bytes": 594582001,
-      "data_revision": "578593ef28ee7a2d861ebf38162c928291b0d4a9"
+      "data_revision": "578593ef28ee7a2d861ebf38162c928291b0d4a9",
+      "size_scope": "entire subset directory, including documentation and integrity files"
     }
   ],
   "planned_subsets": [],
-  "data_revision": "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
+  "data_revision": "ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e",
   "publication_authorization": "All data were authorized before public release.",
-  "pretty_name": "RoboValue Training and Evaluation Data",
-  "training_data_revision": "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
-  "revision_note": "data_revision and training_data_revision refer to existing training payloads; use the commit containing both public test subsets for evaluation/real-world/test-30 and evaluation/simulation/test-30.",
-  "evaluation_data_revision": "578593ef28ee7a2d861ebf38162c928291b0d4a9"
+  "pretty_name": "RoboValue Training and Small-Scale Evaluation Data",
+  "training_data_revision": "ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e",
+  "revision_note": "training_data_revision identifies the unchanged training payloads under training-data/. source_training_data_revision records the original release under data/. evaluation_data_revision identifies the separate public test payloads.",
+  "evaluation_data_revision": "578593ef28ee7a2d861ebf38162c928291b0d4a9",
+  "source_training_data_revision": "7fd21cb2519b4d59cbe9fe8b23cfdf477e3ad145",
+  "training_path_revision": "ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e",
+  "task_alignment": {
+    "documentation": {
+      "en": "TASK_ALIGNMENT.md",
+      "zh": "TASK_ALIGNMENT_zh.md"
+    },
+    "crosswalk": "task_alignment.json",
+    "reference_pdf_sha256": "5c9301aaa774a92ccb55832a1b8f1b6deeb93e77c7c450c79419a8c7bae5cfc8",
+    "status": "passed_with_documented_differences",
+    "training_tasks": 35,
+    "shared_test_tasks": 8
+  }
 }

task_alignment.json

@@ -0,0 +1,1179 @@
+{
+  "schema_version": 1,
+  "status": "passed_with_documented_differences",
+  "paper": {
+    "title": "RoboValue: A Fine-Grained Sim-and-Real Benchmark for Unified Evaluation of Robotic Value Models",
+    "url": "https://arxiv.org/abs/2610.26663",
+    "audited_pdf_sha256": "5c9301aaa774a92ccb55832a1b8f1b6deeb93e77c7c450c79419a8c7bae5cfc8",
+    "audited_pdf_page_count": 89,
+    "page_numbering": "1-based PDF pages equal printed page numbers in the cited pages."
+  },
+  "audited_repo_revision": "7c3e8c7f730ac1971b93cdd9efb444a22da81cb5",
+  "scope": "All four training task vocabularies and native feature schemas; all publicly available simulation One-Shot subtask labels; both public test-30 task/semantic annotation tables; the supplied PDF task cards.",
+  "summary": {
+    "training_tasks": 35,
+    "simulation_training_tasks": 15,
+    "real_world_training_tasks": 20,
+    "all_training_task_names_correspond_to_paper": true,
+    "same_domain_full_shot_and_one_shot_instructions_identical": true,
+    "shared_public_test_tasks": 8,
+    "shared_test_instructions_equal_paper": 8,
+    "shared_test_instructions_equal_training": 7,
+    "simulation_supplemental_training_definitions_exactly_equal_paper": 58,
+    "simulation_supplemental_segments": 58,
+    "simulation_supplemental_frame_labels": 8080,
+    "nonempty_public_test_definitions": 30,
+    "nonempty_public_test_definitions_exactly_equal_paper": 28,
+    "public_test_definition_wording_differences": 2
+  },
+  "training_annotation_availability": [
+    {
+      "domain": "real-world",
+      "setting": "full-shot",
+      "root": "training-data/real-world/full-shot",
+      "native_subtask_feature": false,
+      "public_subtask_annotations": false
+    },
+    {
+      "domain": "real-world",
+      "setting": "one-shot",
+      "root": "training-data/real-world/one-shot",
+      "native_subtask_feature": false,
+      "public_subtask_annotations": false
+    },
+    {
+      "domain": "simulation",
+      "setting": "full-shot",
+      "root": "training-data/simulation/full-shot",
+      "native_subtask_feature": false,
+      "public_subtask_annotations": false
+    },
+    {
+      "domain": "simulation",
+      "setting": "one-shot",
+      "root": "training-data/simulation/one-shot",
+      "native_subtask_feature": false,
+      "public_subtask_annotations": true
+    }
+  ],
+  "annotation_contract": {
+    "paper_section": "3.4",
+    "pdf_page": 7,
+    "native_training_subtask_boundaries": "The original training split does not include subtask boundary annotations. The simulation One-Shot annotations/subtasks supplement is separately constructed for model adaptation.",
+    "boundary_correspondence": "Alignment is semantic. Different training/test trajectories have distinct time axes and interval endpoints.",
+    "csvc": "Execution-order units may group multiple annotated subtasks without changing the evaluation definitions or interval endpoints. CSVC matches semantic identities, not universal local order numbers.",
+    "simulation_semantic_overlay": "Original first/second/third order-slot labels remain separate from the object/action/arm semantic_alignment.json overlay used by CSVC.",
+    "real_world_flower_identity": "Per-episode flower_objects_by_subtask retains flower identity across different valid flower orders."
+  },
+  "training_task_catalog": [
+    {
+      "domain": "simulation",
+      "training_task_index": 0,
+      "task_key": "arrange_largest_number",
+      "task_name": "Arrange Largest Number",
+      "paper_section": "C.2",
+      "pdf_page": 29,
+      "printed_page": 29,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Arrange the numbers from left to right to form the largest possible number, and place them on the pad.",
+      "paper_instruction": "Arrange the numbers from left to right to form the largest possible number, and place them on the pad.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 1,
+      "task_key": "fill_pen_holder",
+      "task_name": "Fill Pen Holder",
+      "paper_section": "C.2",
+      "pdf_page": 29,
+      "printed_page": 29,
+      "paper_subtask_definition_count": 6,
+      "training_instruction": "Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.",
+      "paper_instruction": "Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 2,
+      "task_key": "fold_clothes",
+      "task_name": "Fold Clothes",
+      "paper_section": "C.2",
+      "pdf_page": 30,
+      "printed_page": 30,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Fold the left sleeve inward toward the opposite chest point, fold the right sleeve inward toward the opposite chest point, then fold the lower hem upward until it is aligned with the shoulders.",
+      "paper_instruction": "Fold the left sleeve inward toward the opposite chest point, fold the right sleeve inward toward the opposite chest point, then fold the lower hem upward until it is aligned with the shoulders.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 3,
+      "task_key": "hang_mugs",
+      "task_name": "Hang Mugs",
+      "paper_section": "C.2",
+      "pdf_page": 30,
+      "printed_page": 30,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Hang all three mugs on the mug rack.",
+      "paper_instruction": "Hang all three mugs on the mug rack.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 4,
+      "task_key": "insert_tubes",
+      "task_name": "Insert Tubes",
+      "paper_section": "C.2",
+      "pdf_page": 31,
+      "printed_page": 31,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Insert the three tubes into the rack one by one.",
+      "paper_instruction": "Insert the three tubes into the rack one by one.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 5,
+      "task_key": "make_toast",
+      "task_name": "Make Toast",
+      "paper_section": "C.2",
+      "pdf_page": 31,
+      "printed_page": 31,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Pick up two bread slices one by one, place them into the toaster, and press the lever down.",
+      "paper_instruction": "Pick up two bread slices one by one, place them into the toaster, and press the lever down.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 6,
+      "task_key": "organize_table",
+      "task_name": "Organize Table",
+      "paper_section": "C.2",
+      "pdf_page": 32,
+      "printed_page": 32,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Place the alarm clock on the drawer, put the figurine on its stand, place the mouse on the mouse pad, and push the keyboard into the frame.",
+      "paper_instruction": "Place the alarm clock on the drawer, put the figurine on its stand, place the mouse on the mouse pad, and push the keyboard into the frame.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 7,
+      "task_key": "pack_objects_into_box",
+      "task_name": "Pack Objects Into Box",
+      "paper_section": "C.2",
+      "pdf_page": 33,
+      "printed_page": 33,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Move the box to the center of the table, then place all the objects into the box.",
+      "paper_instruction": "Move the box to the center of the table, then place all the objects into the box.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 8,
+      "task_key": "play_stacking_toy",
+      "task_name": "Play Stacking Toy",
+      "paper_section": "C.2",
+      "pdf_page": 34,
+      "printed_page": 34,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Place the four matching pieces onto the longest peg one by one, place the three matching pieces onto the second-longest peg, place the two matching pieces onto the second-shortest peg, then place the remaining piece onto the shortest peg.",
+      "paper_instruction": "Place the four matching pieces onto the longest peg one by one, place the three matching pieces onto the second-longest peg, place the two matching pieces onto the second-shortest peg, then place the remaining piece onto the shortest peg.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 9,
+      "task_key": "press_by_number",
+      "task_name": "Press By Number",
+      "paper_section": "C.2",
+      "pdf_page": 32,
+      "printed_page": 32,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.",
+      "paper_instruction": "Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 10,
+      "task_key": "put_bottles_into_dustbin",
+      "task_name": "Put Bottles Into Dustbin",
+      "paper_section": "C.2",
+      "pdf_page": 33,
+      "printed_page": 33,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Pick up all four bottles one by one and throw them into the dustbin.",
+      "paper_instruction": "Pick up all four bottles one by one and throw them into the dustbin.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 11,
+      "task_key": "stack_blocks",
+      "task_name": "Stack Blocks",
+      "paper_section": "C.2",
+      "pdf_page": 34,
+      "printed_page": 34,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Stack the three blocks with different textures into one stable pile.",
+      "paper_instruction": "Stack the three blocks with different textures into one stable pile.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 12,
+      "task_key": "store_laptop_and_headphones",
+      "task_name": "Store Laptop And Headphones",
+      "paper_section": "C.2",
+      "pdf_page": 35,
+      "printed_page": 35,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Hang the headphones on the headphone stand, close the laptop, then place the closed laptop into the vertical laptop stand.",
+      "paper_instruction": "Hang the headphones on the headphone stand, close the laptop, then place the closed laptop into the vertical laptop stand.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 13,
+      "task_key": "swap_blocks",
+      "task_name": "Swap Blocks",
+      "paper_section": "C.2",
+      "pdf_page": 35,
+      "printed_page": 35,
+      "paper_subtask_definition_count": 6,
+      "training_instruction": "Swap the two blocks using the empty mat, pressing the button after each move.",
+      "paper_instruction": "Swap the two blocks using the empty mat, pressing the button after each move.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "simulation",
+      "training_task_index": 14,
+      "task_key": "sweep_blocks",
+      "task_name": "Sweep Blocks",
+      "paper_section": "C.2",
+      "pdf_page": 36,
+      "printed_page": 36,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Pick up the broom, pick up and place the dustpan upright on the table, then use the broom to sweep all the objects into the dustpan.",
+      "paper_instruction": "Pick up the broom, pick up and place the dustpan upright on the table, then use the broom to sweep all the objects into the dustpan.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 0,
+      "task_key": "transfer_blocks_using_plate",
+      "task_name": "Transfer Blocks Using Plate",
+      "paper_section": "D.2",
+      "pdf_page": 49,
+      "printed_page": 49,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Place the three colored blocks on the plate one at a time in the order red, green, then blue. Tilt the loaded plate to pour all three blocks into the box.",
+      "paper_instruction": "Place the three colored blocks on the plate one at a time in the order red, green, then blue. Tilt the loaded plate to pour all three blocks into the box.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 1,
+      "task_key": "enter_keypad_number_sequence",
+      "task_name": "Enter Keypad Number Sequence",
+      "paper_section": "D.2",
+      "pdf_page": 42,
+      "printed_page": 42,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Press the keys 2, 8, 2, 6, and 8 once each in that order (five presses total), fully releasing each key and returning to the waiting position in front of the keypad after each press.",
+      "paper_instruction": "Press the keys in the sequence 2-8-2-6-8, fully releasing each key and returning to the waiting position in front of the keypad between presses.",
+      "training_instruction_exactly_matches_paper": false,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 2,
+      "task_key": "set_the_table",
+      "task_name": "Set the Table",
+      "paper_section": "D.2",
+      "pdf_page": 48,
+      "printed_page": 48,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Take the plate, bowl, fork, and knife out of the crate. Place the plate face up in the lower center of the placemat and the bowl open side up in the upper center, in front of the plate. Place the fork on the left and the knife on the right, with the fork tines and knife tip pointing toward the front. Complete these actions in any order.",
+      "paper_instruction": "Take the plate, bowl, fork, and knife out of the crate. Place the plate face up in the lower center of the placemat and the bowl open side up in the upper center, in front of the plate. Place the fork on the left and the knife on the right, with the fork tines and knife tip pointing toward the front. Complete these actions in any order.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 3,
+      "task_key": "arrange_seasoning_jars",
+      "task_name": "Arrange Seasoning Jars",
+      "paper_section": "D.2",
+      "pdf_page": 41,
+      "printed_page": 41,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Place the three seasoning jars on the two-tier rack in any order, on either tier and in any positions.",
+      "paper_instruction": "Place the three seasoning jars on the two-tier rack in any order, on either tier and in any positions.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 4,
+      "task_key": "load_plastic_plates_into_rack",
+      "task_name": "Load Plastic Plates Into Rack",
+      "paper_section": "D.2",
+      "pdf_page": 43,
+      "printed_page": 43,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Insert the three plastic plates into the rack in any order and in any slots.",
+      "paper_instruction": "Insert the three plastic plates into the rack in any order and in any slots.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 5,
+      "task_key": "press_button_five_times",
+      "task_name": "Press Button Five Times",
+      "paper_section": "D.2",
+      "pdf_page": 47,
+      "printed_page": 47,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Press the blue button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "paper_instruction": "Press the button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "training_instruction_exactly_matches_paper": false,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 6,
+      "task_key": "place_pot_on_induction_cooker",
+      "task_name": "Place Pot On Induction Cooker",
+      "paper_section": "D.2",
+      "pdf_page": 46,
+      "printed_page": 46,
+      "paper_subtask_definition_count": 4,
+      "training_instruction": "Place the pot, with the spatula inside, on the induction cooker so that most of the pot is over the cooking surface. Turn the handle toward the camera, move the spatula to the basket, and cover the pot with the lid.",
+      "paper_instruction": "Place the pot, with the spatula inside, on the induction cooker so that most of the pot is over the cooking surface. Turn the handle toward the camera, move the spatula to the basket, and cover the pot with the lid.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 7,
+      "task_key": "organize_desktop_workstation",
+      "task_name": "Organize Desktop Workstation",
+      "paper_section": "D.2",
+      "pdf_page": 45,
+      "printed_page": 45,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.",
+      "paper_instruction": "Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 8,
+      "task_key": "organize_books_and_pen",
+      "task_name": "Organize Books And Pen",
+      "paper_section": "D.2",
+      "pdf_page": 44,
+      "printed_page": 44,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Center and align the closed book in the target rectangle. Place the pen along the spine of the open book, close the book, and stack it on the first book.",
+      "paper_instruction": "Center and align the closed book in the target rectangle. Place the pen along the spine of the open book, close the book, and stack it on the first book.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 9,
+      "task_key": "arrange_family_tea_set",
+      "task_name": "Arrange Family Tea Set",
+      "paper_section": "D.2",
+      "pdf_page": 40,
+      "printed_page": 40,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Place the teacup on the cat-head mat, the teapot on the yellow mat, the lid on the teapot, the tongs in their holder, and the canister on the round mat. Complete these placements in any order, provided the teapot is placed before its lid.",
+      "paper_instruction": "Place the teacup on the cat-head mat, the teapot on the yellow mat, the lid on the teapot, the tongs in their holder, and the canister on the round mat. Complete these placements in any order, provided the teapot is placed before its lid.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 10,
+      "task_key": "operate_drawers_in_sequence",
+      "task_name": "Operate Drawers In Sequence",
+      "paper_section": "D.2",
+      "pdf_page": 43,
+      "printed_page": 43,
+      "paper_subtask_definition_count": 6,
+      "training_instruction": "Open the bottom, middle, and top drawers in that order, then close the top, middle, and bottom drawers in that order.",
+      "paper_instruction": "Open the bottom, middle, and top drawers in that order, then close the top, middle, and bottom drawers in that order.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 11,
+      "task_key": "sort_blocks_in_order",
+      "task_name": "Sort Blocks in Order",
+      "paper_section": "D.2",
+      "pdf_page": 48,
+      "printed_page": 48,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Place the red block on the yellow drawer, the yellow block on the white drawer, and the blue block on the dark-green drawer, in that order.",
+      "paper_instruction": "Place the red block on the yellow drawer, the yellow block on the white drawer, and the blue block on the dark-green drawer, in that order.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 12,
+      "task_key": "fill_and_place_pen_holder",
+      "task_name": "Fill And Place Pen Holder",
+      "paper_section": "D.2",
+      "pdf_page": 42,
+      "printed_page": 42,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Hold the pen holder with its opening upward, insert all four pens one by one in any order, then place the filled holder upright.",
+      "paper_instruction": "Hold the pen holder with its opening upward, insert all four pens one by one in any order, then place the filled holder upright.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 13,
+      "task_key": "organize_cosmetics_storage_box",
+      "task_name": "Organize Cosmetics Storage Box",
+      "paper_section": "D.2",
+      "pdf_page": 44,
+      "printed_page": 44,
+      "paper_subtask_definition_count": 7,
+      "training_instruction": "Open the upper drawer, cap the lipstick, and place the lipstick and eyebrow pencil inside. Close the drawer once both are inside. Along the way, place the serum and cleansing oil in the storage area above, in either order, before or after closing the drawer.",
+      "paper_instruction": "Open the upper drawer, cap the lipstick, and place the lipstick and eyebrow pencil inside. Close the drawer once both are inside. Place the serum and cleansing oil in the storage area above, in either order, before or after closing the drawer.",
+      "training_instruction_exactly_matches_paper": false,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 14,
+      "task_key": "strike_drum_five_times",
+      "task_name": "Strike Drum Five Times",
+      "paper_section": "D.2",
+      "pdf_page": 49,
+      "printed_page": 49,
+      "paper_subtask_definition_count": 6,
+      "training_instruction": "Pick up the drumstick and strike the center of the drumhead 5 times continuously.",
+      "paper_instruction": "Pick up the drumstick and strike the center of the drumhead 5 times continuously.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 15,
+      "task_key": "apply_dish_soap_repeatedly",
+      "task_name": "Apply Dish Soap Repeatedly",
+      "paper_section": "D.2",
+      "pdf_page": 40,
+      "printed_page": 40,
+      "paper_subtask_definition_count": 7,
+      "training_instruction": "Press the dish-soap bottle 5 times continuously, dispensing soap onto the dishcloth.",
+      "paper_instruction": "Press the dish-soap bottle 5 times continuously, dispensing soap onto the dishcloth.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 16,
+      "task_key": "season_meat_skewers_repeatedly",
+      "task_name": "Season Meat Skewers Repeatedly",
+      "paper_section": "D.2",
+      "pdf_page": 47,
+      "printed_page": 47,
+      "paper_subtask_definition_count": 6,
+      "training_instruction": "Sprinkle the meat skewers with the seasoning shaker 5 times continuously.",
+      "paper_instruction": "Sprinkle the meat skewers with the seasoning shaker 5 times continuously.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 17,
+      "task_key": "pack_items_for_shipping",
+      "task_name": "Pack Items For Shipping",
+      "paper_section": "D.2",
+      "pdf_page": 45,
+      "printed_page": 45,
+      "paper_subtask_definition_count": 3,
+      "training_instruction": "Place the wireless mouse, then the wired earphones, in the cardboard box. Cover both items with bubble wrap.",
+      "paper_instruction": "Place the wireless mouse, then the wired earphones, in the cardboard box. Cover both items with bubble wrap.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 18,
+      "task_key": "arrange_flowers_and_reed_diffuser",
+      "task_name": "Arrange Flowers and Reed Diffuser",
+      "paper_section": "D.2",
+      "pdf_page": 41,
+      "printed_page": 41,
+      "paper_subtask_definition_count": 5,
+      "training_instruction": "Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.",
+      "paper_instruction": "Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": true
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 19,
+      "task_key": "prepare_breakfast",
+      "task_name": "Prepare Breakfast",
+      "paper_section": "D.2",
+      "pdf_page": 46,
+      "printed_page": 46,
+      "paper_subtask_definition_count": 6,
+      "training_instruction": "Place the pig bun on the left, the yellow steamed bun on the right, and the black steamed bun at the front of the upper tier of the two already assembled, in that order. Add the empty tier on top, place the large bun in its center, then cover the steamer with the lid.",
+      "paper_instruction": "Place the pig bun on the left, the yellow steamed bun on the right, and the black steamed bun at the front of the upper tier of the two already assembled, in that order. Add the empty tier on top, place the large bun in its center, then cover the steamer with the lid.",
+      "training_instruction_exactly_matches_paper": true,
+      "shared_with_public_test_30": false
+    }
+  ],
+  "shared_task_alignment": [
+    {
+      "domain": "simulation",
+      "task_key": "fill_pen_holder",
+      "task_name": "Fill Pen Holder",
+      "training_task_index": 1,
+      "public_test_task_index": 0,
+      "pdf_page": 29,
+      "printed_page": 29,
+      "training_instruction": "Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.",
+      "public_test_instruction": "Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.",
+      "paper_instruction": "Pick up the pen holder, place all the pens into it one by one, then put the filled holder back on the table.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": false,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Hold the empty pen holder.",
+          "training_supplemental_id": "S1",
+          "training_supplemental_definition_en": "Hold the empty pen holder.",
+          "public_test_id": "001",
+          "public_test_definition_en": "Hold the empty pen holder.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Insert the first selected pen into the pen holder.",
+          "training_supplemental_id": "S2",
+          "training_supplemental_definition_en": "Insert the first selected pen into the pen holder.",
+          "public_test_id": "002",
+          "public_test_definition_en": "Insert the first selected pen into the pen holder.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Insert the second selected pen into the pen holder.",
+          "training_supplemental_id": "S3",
+          "training_supplemental_definition_en": "Insert the second selected pen into the pen holder.",
+          "public_test_id": "003",
+          "public_test_definition_en": "Insert the second selected pen into the pen holder.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Insert the third selected pen into the pen holder.",
+          "training_supplemental_id": "S4",
+          "training_supplemental_definition_en": "Insert the third selected pen into the pen holder.",
+          "public_test_id": "004",
+          "public_test_definition_en": "Insert the third selected pen into the pen holder.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S5",
+          "paper_definition_en": "Insert the remaining pen into the pen holder.",
+          "training_supplemental_id": "S5",
+          "training_supplemental_definition_en": "Insert the remaining pen into the pen holder.",
+          "public_test_id": "005",
+          "public_test_definition_en": "Insert the remaining pen into the pen holder.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S6",
+          "paper_definition_en": "Place the filled pen holder upright on the table.",
+          "training_supplemental_id": "S6",
+          "training_supplemental_definition_en": "Place the filled pen holder upright on the table.",
+          "public_test_id": "006",
+          "public_test_definition_en": "Place the filled pen holder upright on the table.",
+          "test_definition_exactly_matches_paper": true
+        }
+      ],
+      "paper_order_reference": "The pens may be inserted in any order. Picking up the holder precedes insertion, and putting it down follows all four insertions.",
+      "training_subtask_boundary_comparison": "Supplemental training and test trajectories are distinct executions; time boundaries are not required to be numerically equal."
+    },
+    {
+      "domain": "simulation",
+      "task_key": "hang_mugs",
+      "task_name": "Hang Mugs",
+      "training_task_index": 3,
+      "public_test_task_index": 1,
+      "pdf_page": 30,
+      "printed_page": 30,
+      "training_instruction": "Hang all three mugs on the mug rack.",
+      "public_test_instruction": "Hang all three mugs on the mug rack.",
+      "paper_instruction": "Hang all three mugs on the mug rack.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": false,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Hang the first cup on an empty hook.",
+          "training_supplemental_id": "S1",
+          "training_supplemental_definition_en": "Hang the first cup on an empty hook.",
+          "public_test_id": "001",
+          "public_test_definition_en": "Hang the first cup on an empty hook.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Hang the second cup on another empty hook.",
+          "training_supplemental_id": "S2",
+          "training_supplemental_definition_en": "Hang the second cup on another empty hook.",
+          "public_test_id": "002",
+          "public_test_definition_en": "Hang the second cup on another empty hook.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Hang the third cup on the remaining empty hook.",
+          "training_supplemental_id": "S3",
+          "training_supplemental_definition_en": "Hang the third cup on the remaining empty hook.",
+          "public_test_id": "003",
+          "public_test_definition_en": "Hang the third cup on the remaining empty hook.",
+          "test_definition_exactly_matches_paper": true
+        }
+      ],
+      "paper_order_reference": "The mugs may be hung in any order.",
+      "training_subtask_boundary_comparison": "Supplemental training and test trajectories are distinct executions; time boundaries are not required to be numerically equal."
+    },
+    {
+      "domain": "simulation",
+      "task_key": "press_by_number",
+      "task_name": "Press By Number",
+      "training_task_index": 9,
+      "public_test_task_index": 2,
+      "pdf_page": 32,
+      "printed_page": 32,
+      "training_instruction": "Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.",
+      "public_test_instruction": "Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.",
+      "paper_instruction": "Press the two red buttons the required number of times according to the number cards, then press the blue button to confirm.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": true,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Press the first button the required number of times.",
+          "training_supplemental_id": "S1",
+          "training_supplemental_definition_en": "Press the first button the required number of times.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Press the blue button.",
+          "training_supplemental_id": "S2",
+          "training_supplemental_definition_en": "Press the blue button.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Press the second button the required number of times.",
+          "training_supplemental_id": "S3",
+          "training_supplemental_definition_en": "Press the second button the required number of times.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Press the blue button again.",
+          "training_supplemental_id": "S4",
+          "training_supplemental_definition_en": "Press the blue button again.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        }
+      ],
+      "paper_order_reference": "The required press counts vary with the cards. Complete the redbutton presses before pressing the blue confirmation button.",
+      "training_subtask_boundary_comparison": "Supplemental training and test trajectories are distinct executions; time boundaries are not required to be numerically equal."
+    },
+    {
+      "domain": "simulation",
+      "task_key": "swap_blocks",
+      "task_name": "Swap Blocks",
+      "training_task_index": 13,
+      "public_test_task_index": 3,
+      "pdf_page": 35,
+      "printed_page": 35,
+      "training_instruction": "Swap the two blocks using the empty mat, pressing the button after each move.",
+      "public_test_instruction": "Swap the two blocks using the empty mat, pressing the button after each move.",
+      "paper_instruction": "Swap the two blocks using the empty mat, pressing the button after each move.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": true,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Put the first block on the empty mat.",
+          "training_supplemental_id": "S1",
+          "training_supplemental_definition_en": "Put the first block on the empty mat.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Press the button to confirm the first move.",
+          "training_supplemental_id": "S2",
+          "training_supplemental_definition_en": "Press the button to confirm the first move.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Transfer the second block onto the original first block’s position.",
+          "training_supplemental_id": "S3",
+          "training_supplemental_definition_en": "Transfer the second block onto the original first block’s position.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Press the button to confirm the second move.",
+          "training_supplemental_id": "S4",
+          "training_supplemental_definition_en": "Press the button to confirm the second move.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S5",
+          "paper_definition_en": "Put the first block on the original second block’s mat.",
+          "training_supplemental_id": "S5",
+          "training_supplemental_definition_en": "Put the first block on the original second block’s mat.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        },
+        {
+          "paper_id": "S6",
+          "paper_definition_en": "Press the button to confirm the last move.",
+          "training_supplemental_id": "S6",
+          "training_supplemental_definition_en": "Press the button to confirm the last move.",
+          "public_test_id": null,
+          "public_test_definition_en": null,
+          "test_definition_exactly_matches_paper": null
+        }
+      ],
+      "paper_order_reference": "Press the button after every block move.",
+      "training_subtask_boundary_comparison": "Supplemental training and test trajectories are distinct executions; time boundaries are not required to be numerically equal."
+    },
+    {
+      "domain": "real-world",
+      "task_key": "organize_desktop_workstation",
+      "task_name": "Organize Desktop Workstation",
+      "training_task_index": 7,
+      "public_test_task_index": 0,
+      "pdf_page": 45,
+      "printed_page": 45,
+      "training_instruction": "Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.",
+      "public_test_instruction": "Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.",
+      "paper_instruction": "Place the pen in the holder, place the eraser in the open drawer and close it, switch off the lamp, and unplug it. These actions may be performed in any order, except that the lamp must be switched off before it is unplugged.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": false,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Place the pen in the pen holder.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "001",
+          "public_test_definition_en": "Place the pen in the pen holder.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Place the eraser in the drawer.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "002",
+          "public_test_definition_en": "Place the eraser in the storage box.",
+          "test_definition_exactly_matches_paper": false
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Close the drawer.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "003",
+          "public_test_definition_en": "Close the storage box.",
+          "test_definition_exactly_matches_paper": false
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Turn off the desk lamp.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "004",
+          "public_test_definition_en": "Turn off the desk lamp.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S5",
+          "paper_definition_en": "Unplug the desk lamp.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "005",
+          "public_test_definition_en": "Unplug the desk lamp.",
+          "test_definition_exactly_matches_paper": true
+        }
+      ],
+      "paper_order_reference": null,
+      "training_subtask_boundary_comparison": "No public training subtask boundaries are available for this task."
+    },
+    {
+      "domain": "real-world",
+      "task_key": "arrange_flowers_and_reed_diffuser",
+      "task_name": "Arrange Flowers and Reed Diffuser",
+      "training_task_index": 18,
+      "public_test_task_index": 1,
+      "pdf_page": 41,
+      "printed_page": 41,
+      "training_instruction": "Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.",
+      "public_test_instruction": "Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.",
+      "paper_instruction": "Place the pink, white, and yellow flowers consecutively in the vase, in any order. Place the diffuser bottle on the coaster and insert the reed stick into it, in either order. Either diffuser action may occur before or after the flower sequence, but neither may occur between flower placements.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": false,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Insert the first flower into the vase.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "001",
+          "public_test_definition_en": "Insert the first flower into the vase.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Insert the second flower into the vase.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "002",
+          "public_test_definition_en": "Insert the second flower into the vase.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Insert the third flower into the vase.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "003",
+          "public_test_definition_en": "Insert the third flower into the vase.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Place the reed diffuser bottle on the coaster.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "004",
+          "public_test_definition_en": "Place the reed diffuser bottle on the coaster.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S5",
+          "paper_definition_en": "Insert the diffuser reed into the reed diffuser bottle.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "005",
+          "public_test_definition_en": "Insert the diffuser reed into the reed diffuser bottle.",
+          "test_definition_exactly_matches_paper": true
+        }
+      ],
+      "paper_order_reference": null,
+      "training_subtask_boundary_comparison": "No public training subtask boundaries are available for this task."
+    },
+    {
+      "domain": "real-world",
+      "task_key": "strike_drum_five_times",
+      "task_name": "Strike Drum Five Times",
+      "training_task_index": 14,
+      "public_test_task_index": 2,
+      "pdf_page": 49,
+      "printed_page": 49,
+      "training_instruction": "Pick up the drumstick and strike the center of the drumhead 5 times continuously.",
+      "public_test_instruction": "Pick up the drumstick and strike the center of the drumhead 5 times continuously.",
+      "paper_instruction": "Pick up the drumstick and strike the center of the drumhead 5 times continuously.",
+      "instruction_equal_training_test": true,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": false,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Strike the center of the drum for the first time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "001",
+          "public_test_definition_en": "Strike the center of the drum for the first time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Strike the center of the drum for the second time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "002",
+          "public_test_definition_en": "Strike the center of the drum for the second time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Strike the center of the drum for the third time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "003",
+          "public_test_definition_en": "Strike the center of the drum for the third time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Strike the center of the drum for the fourth time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "004",
+          "public_test_definition_en": "Strike the center of the drum for the fourth time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S5",
+          "paper_definition_en": "Strike the center of the drum for the fifth time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "005",
+          "public_test_definition_en": "Strike the center of the drum for the fifth time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S6",
+          "paper_definition_en": "Place the drumstick on the table.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "006",
+          "public_test_definition_en": "Place the drumstick on the table.",
+          "test_definition_exactly_matches_paper": true
+        }
+      ],
+      "paper_order_reference": null,
+      "training_subtask_boundary_comparison": "No public training subtask boundaries are available for this task."
+    },
+    {
+      "domain": "real-world",
+      "task_key": "press_button_five_times",
+      "task_name": "Press Button Five Times",
+      "training_task_index": 5,
+      "public_test_task_index": 3,
+      "pdf_page": 47,
+      "printed_page": 47,
+      "training_instruction": "Press the blue button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "public_test_instruction": "Press the button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "paper_instruction": "Press the button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "instruction_equal_training_test": false,
+      "instruction_equal_test_paper": true,
+      "public_test_subtask_table_empty": false,
+      "subtask_definition_comparison": [
+        {
+          "paper_id": "S1",
+          "paper_definition_en": "Press the button for the first time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "001",
+          "public_test_definition_en": "Press the button for the first time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S2",
+          "paper_definition_en": "Press the button for the second time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "002",
+          "public_test_definition_en": "Press the button for the second time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S3",
+          "paper_definition_en": "Press the button for the third time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "003",
+          "public_test_definition_en": "Press the button for the third time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S4",
+          "paper_definition_en": "Press the button for the fourth time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "004",
+          "public_test_definition_en": "Press the button for the fourth time.",
+          "test_definition_exactly_matches_paper": true
+        },
+        {
+          "paper_id": "S5",
+          "paper_definition_en": "Press the button for the fifth time.",
+          "training_supplemental_id": null,
+          "training_supplemental_definition_en": null,
+          "public_test_id": "005",
+          "public_test_definition_en": "Press the button for the fifth time.",
+          "test_definition_exactly_matches_paper": true
+        }
+      ],
+      "paper_order_reference": null,
+      "training_subtask_boundary_comparison": "No public training subtask boundaries are available for this task."
+    }
+  ],
+  "wording_and_coverage_differences": [
+    {
+      "id": "desktop_object_wording",
+      "domain": "real-world",
+      "task_name": "Organize Desktop Workstation",
+      "pdf_page": 45,
+      "public_test_ids": [
+        "002",
+        "003"
+      ],
+      "public_test_definitions": [
+        "Place the eraser in the storage box.",
+        "Close the storage box."
+      ],
+      "paper_definitions": [
+        "Place the eraser in the drawer.",
+        "Close the drawer."
+      ],
+      "status": "documented_wording_difference",
+      "resolution": "The comparison retains both source wordings; original labels and interval endpoints remain unchanged. Canonical paper wording may be read from this alignment index."
+    },
+    {
+      "id": "button_color_qualifier",
+      "domain": "real-world",
+      "task_name": "Press Button Five Times",
+      "pdf_page": 47,
+      "training": "Press the blue button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "public_test_and_paper": "Press the button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "status": "wording_variant",
+      "resolution": "Training adds a blue-color qualifier. Both specify five presses and full release/return; no count change."
+    },
+    {
+      "id": "simulation_mem_coverage",
+      "domain": "simulation",
+      "task_names": [
+        "Press By Number",
+        "Swap Blocks"
+      ],
+      "pdf_pages": [
+        32,
+        35
+      ],
+      "training_and_paper_definition_counts": [
+        4,
+        6
+      ],
+      "public_test_definition_counts": [
+        0,
+        0
+      ],
+      "status": "coverage_difference",
+      "resolution": "MEM task definitions and interval tables remain empty in the public test source. Paper/training descriptions are reference definitions, not inferred counting ground truth."
+    },
+    {
+      "id": "press_by_number_paper_internal_conflict",
+      "domain": "simulation",
+      "task_name": "Press By Number",
+      "pdf_page": 32,
+      "status": "requires_author_protocol_clarification",
+      "instruction_procedure_and_order": "Complete both red-button press counts before blue confirmation.",
+      "printed_subtask_sequence": "First red-button count, blue confirmation, second red-button count, blue confirmation again.",
+      "resolution": "The card is internally inconsistent about the confirmation count/order. Existing training taxonomy faithfully follows the printed S1-S4 list and instructions follow the printed instruction. Preserve labels; author protocol clarification is required before treating them as one unambiguous sequence."
+    }
+  ],
+  "other_training_instruction_wording_variants": [
+    {
+      "domain": "real-world",
+      "training_task_index": 1,
+      "task_name": "Enter Keypad Number Sequence",
+      "training_instruction": "Press the keys 2, 8, 2, 6, and 8 once each in that order (five presses total), fully releasing each key and returning to the waiting position in front of the keypad after each press.",
+      "paper_instruction": "Press the keys in the sequence 2-8-2-6-8, fully releasing each key and returning to the waiting position in front of the keypad between presses.",
+      "pdf_page": 42
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 5,
+      "task_name": "Press Button Five Times",
+      "training_instruction": "Press the blue button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "paper_instruction": "Press the button 5 times, fully releasing it and returning to the waiting position after each press.",
+      "pdf_page": 47
+    },
+    {
+      "domain": "real-world",
+      "training_task_index": 13,
+      "task_name": "Organize Cosmetics Storage Box",
+      "training_instruction": "Open the upper drawer, cap the lipstick, and place the lipstick and eyebrow pencil inside. Close the drawer once both are inside. Along the way, place the serum and cleansing oil in the storage area above, in either order, before or after closing the drawer.",
+      "paper_instruction": "Open the upper drawer, cap the lipstick, and place the lipstick and eyebrow pencil inside. Close the drawer once both are inside. Place the serum and cleansing oil in the storage area above, in either order, before or after closing the drawer.",
+      "pdf_page": 44
+    }
+  ],
+  "metric_scope": {
+    "pdf_page": 26,
+    "paper_table": "Table 7",
+    "sia_and_csvc_trajectory_types": [
+      "ED",
+      "MS"
+    ],
+    "memory_voc_trajectory_type": "MEM",
+    "fpl_trajectory_types": [
+      "EC",
+      "ER"
+    ],
+    "public_release": "These 60 trajectories are partial diagnostic coverage. Simulation keeps 10 formal metric entries, excluding a formal FPL cohort while retaining original FPL fields; real-world keeps 11."
+  },
+  "data_preservation": "This alignment index does not modify native training payloads or public test annotations, videos, frame indices, timestamps, interval endpoints, object assignments, or arm roles."
+}

{data → training-data}/real-world/full-shot/README.md

@@ -19,9 +19,9 @@ Expert demonstrations collected with the AgiBot Genie02 robot.
 | RGB cameras | Head: 640 × 400; left/right wrist: 640 × 528 each |
 | Camera videos | 6,000 · HEVC |
 
-This directory is an independent LeRobot dataset root. Set `root` to `data/real-world/full-shot` inside the downloaded snapshot. The [download and training examples](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data#download-and-reading) use LeRobot **0.4.2** and PyAV **15.1.0** and return states, actions, task instructions and three camera observations as PyTorch batches. Value or progress supervision is constructed by the training method.
+This directory is an independent LeRobot dataset root. Set `root` to `training-data/real-world/full-shot` inside the downloaded snapshot. The [download and training examples](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data#download-and-reading) use LeRobot **0.4.2** and PyAV **15.1.0** and return states, actions, task instructions and three camera observations as PyTorch batches. Value or progress supervision is constructed by the training method.
 
-[Feature definitions](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/data/real-world/full-shot/meta/info.json) · [Shared trajectory contents and task descriptions](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data#real-world)
+[Feature definitions](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/training-data/real-world/full-shot/meta/info.json) · [Shared trajectory contents and task descriptions](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data#real-world)
 
 <details>
 <summary><strong>Per-task episode allocation</strong></summary>
@@ -55,7 +55,7 @@ Task indices match `meta/tasks.parquet`. Source episode identifiers and task cor
 
 **Publication authorization:** All data were authorized before public release.
 
-**Data revision:** `3bcf64ae6d2a39e0795f413b65bfd08a3a8c6f95`. Use this fixed revision with the download examples.
+**Data revision:** `ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e`. Use this fixed revision with the download examples.
 
 ## 中文
 
@@ -65,10 +65,10 @@ Task indices match `meta/tasks.parquet`. Source episode identifiers and task cor
 
 平台为 **AgiBot Genie02**，采用 **LeRobot Dataset v3.0**，采样率为 **30 FPS**。状态为 `float32[290]`，动作为 `float32[52]`；三路 RGB 相机为头部：640 × 400；左腕、右腕：均为 640 × 528，共 **6,000 个原编码视频（HEVC）**。
 
-下载后，将 LeRobot 的 `root` 设置为快照中的 `data/real-world/full-shot`。主卡片的[下载与训练示例](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/README_zh.md#download-and-reading)提供状态、动作、任务指令与三路相机观测的 PyTorch batch；value 或 progress 监督由训练方法构造。
+下载后，将 LeRobot 的 `root` 设置为快照中的 `training-data/real-world/full-shot`。主卡片的[下载与训练示例](https://huggingface.co/datasets/wzlllong/value-model-benchmark-training-data/blob/main/README_zh.md#download-and-reading)提供状态、动作、任务指令与三路相机观测的 PyTorch batch；value 或 progress 监督由训练方法构造。
 
 折叠表列出任务索引、轨迹分配及数据行范围；原始来源标识与对应关系保留在 `provenance/`。
 
-**数据版本：**`3bcf64ae6d2a39e0795f413b65bfd08a3a8c6f95`，下载与读取时使用该固定版本。
+**数据版本：**`ec8c2f2ae736bae8f19ffd9e0ca5cd16dbbe236e`，下载与读取时使用该固定版本。
 
 **发布授权：**全部数据均经授权后公开发布。

{data → training-data}/real-world/full-shot/checksums.sha256

@@ -1998,10 +1998,10 @@ bd535062dbe63ba098e1f52a744a48739dcf233de1bce12d1eaf9a1212d88f25  data/chunk-001
 ad6956e579b020509c9582bb2a15cbe1c7b39803789f77a2a51caae28a4150c9  data/chunk-001/file-997.parquet
 0ebcdbd1b72526cdb240f78ae757623847d6dd86d264932e7a8d26adb4644892  data/chunk-001/file-998.parquet
 994b33933358cfabdaf2def445762bb06e9e941e0d865188b2e6980168d901e9  data/chunk-001/file-999.parquet
-a1c5174b88db6a6b65fc5f42d6c4888555c9cb344ea44bc3f6cd1c65ace34295  meta/episodes/chunk-000/file-000.parquet
+c5c40c5f86263f440c34c7374a29211bfd8fbc64a7c946a7fad96c0459495c38  meta/episodes/chunk-000/file-000.parquet
 faaf8dbabeb5222b1fdffab5e3d6122ed52417e2435e40ef3ba9fe1d4526c2f7  meta/info.json
 ba82dd5d05ac9583522d84ebe0483659ae9d3b2abc6310c4df4d4fe2e7a97445  meta/stats.json
-fe80e4c17ad7b530ffc48b089fe461d1dbb11d77b5507d0c921832f3d337d855  meta/tasks.parquet
+2610f566973107a58bcf8dd2876f97fd49020d815ddba54262060f372bf3e023  meta/tasks.parquet
 f859e3c477b87aee893f5dc59818a44816d2b640096eb548c148d4af5a68f1bb  provenance/conversion.json
 46f0481b6b00eb8e512bedf9ab6c5b0836582a24cbfc31069663e456c3e1d20c  provenance/episode_mapping.json
 35bed3d2a6cfc18c8ce0787664ab79b361d8521a883b8ea40f22a8ea32e9d764  provenance/episodes.parquet