Fix evaluation code and source links
Browse filesCorrect code links after the src layout migration. Preserve their exact source commits. Link Search attribution to the notices included in this dataset. Task definitions, references and manifests remain unchanged.
README.md
CHANGED
|
@@ -76,7 +76,7 @@ query = task["query"]
|
|
| 76 |
answer, supporting_urls = task["reference_answer"], task["reference_urls"]
|
| 77 |
```
|
| 78 |
|
| 79 |
-
`definitions.tar.gz` contains the same definition bytes as `definitions/`. Browser Eval's [data lock](https://github.com/lexmount/browser-eval/blob/d3a474c85b0c0249b11edf3e712bcd6284d9b008/harness/benchmarks/data-lock.json) selects the installed snapshot. Follow the [Web Search guide](https://github.com/lexmount/browser-eval/blob/d3a474c85b0c0249b11edf3e712bcd6284d9b008/harness/web_search/README.md) to declare a profile, collect provider responses and build a report with saved evidence. The certified collector is `scripts/run_certified_benchmark.py web-search`.
|
| 80 |
|
| 81 |
Reference Hit@k tests whether a supplied supporting URL appears in the first k results; MRR@k measures the first such hit's reciprocal rank. Answer Match@k checks answer cues in returned text. Source references are not an exhaustive set of relevant pages, and lexical answer matches do not establish that a result correctly answers the question. Report these measures separately from graded relevance. Searchable copies of the dataset can also create contamination; Browser Eval exposes a separate contamination diagnostic while retaining the original results.
|
| 82 |
|
|
|
|
| 76 |
answer, supporting_urls = task["reference_answer"], task["reference_urls"]
|
| 77 |
```
|
| 78 |
|
| 79 |
+
`definitions.tar.gz` contains the same definition bytes as `definitions/`. Browser Eval's [data lock](https://github.com/lexmount/browser-eval/blob/d3a474c85b0c0249b11edf3e712bcd6284d9b008/src/harness/benchmarks/data-lock.json) selects the installed snapshot. Follow the [Web Search guide](https://github.com/lexmount/browser-eval/blob/d3a474c85b0c0249b11edf3e712bcd6284d9b008/src/harness/web_search/README.md) to declare a profile, collect provider responses and build a report with saved evidence. The certified collector is `scripts/run_certified_benchmark.py web-search`.
|
| 80 |
|
| 81 |
Reference Hit@k tests whether a supplied supporting URL appears in the first k results; MRR@k measures the first such hit's reciprocal rank. Answer Match@k checks answer cues in returned text. Source references are not an exhaustive set of relevant pages, and lexical answer matches do not establish that a result correctly answers the question. Report these measures separately from graded relevance. Searchable copies of the dataset can also create contamination; Browser Eval exposes a separate contamination diagnostic while retaining the original results.
|
| 82 |
|