Add dataset card

README.md

@@ -1,44 +1,66 @@
 ---
-dataset_info:
-  features:
-  - name: id
-    dtype: string
-  - name: project
-    dtype: string
-  - name: project_rel
-    dtype: string
-  - name: interface_path
-    dtype: string
-  - name: target_path
-    dtype: string
-  - name: interface_code
-    dtype: string
-  - name: prompt_files
-    list:
-    - name: code
-      dtype: string
-    - name: language
-      dtype: string
-    - name: path
-      dtype: string
-  - name: prompt
-    dtype: string
-  - name: answer_prefix
-    dtype: string
-  splits:
-  - name: dev
-    num_bytes: 28512
-    num_examples: 4
-  - name: test
-    num_bytes: 311948
-    num_examples: 30
-  download_size: 154927
-  dataset_size: 340460
-configs:
-- config_name: default
-  data_files:
-  - split: dev
-    path: data/dev-*
-  - split: test
-    path: data/test-*
+pretty_name: API-Upgrade
+license: other
+task_categories:
+- text-generation
+language:
+- en
+tags:
+- code
+- rust
+- software-engineering
+- arxiv:2607.13921
 ---
+
+## API-Upgrade
+
+API-Upgrade evaluates repository-level Rust synthesis against newer dependency APIs.
+
+This dataset was used for evaluation in the paper
+[Generative Compilation: On-the-Fly Compiler Feedback as AI Generates Code](https://arxiv.org/abs/2607.13921).
+You can find the corresponding evaluation code in
+[the project GitHub repository](https://github.com/eth-sri/generative-compilation).
+
+### Example Usage
+
+```python
+from datasets import load_dataset
+import json
+
+dataset = load_dataset("eth-sri/API-Upgrade")
+for instance in dataset["test"]:
+    print(json.dumps(instance, indent=2))
+    break
+```
+
+### Example Instance
+
+```json
+{
+  "id": "test/addr2line__0.25.1_to_0.26.0/output/src/lib.rs",
+  "project": "addr2line__0.25.1_to_0.26.0",
+  "project_rel": "test/addr2line__0.25.1_to_0.26.0/output",
+  "interface_path": "src/lib.rs",
+  "target_path": "src/lib.rs",
+  "interface_code": "pub fn ... {\\n    unimplemented!();\\n}",
+  "prompt_files": [
+    {
+      "path": "Cargo.toml",
+      "code": "...",
+      "language": "toml"
+    }
+  ],
+  "prompt": "...",
+  "answer_prefix": "src/lib.rs\\n```rust\\n"
+}
+```
+
+Each row also includes the complete rendered `prompt`, the `answer_prefix` used
+to start model generation, and the structured source fields needed to inspect
+the task.
+
+### License
+
+The benchmark combines source code from multiple upstream projects. The license
+of each upstream project continues to apply to its code; consult the source files
+and the project repository for the corresponding license notices.