Update README.md
#8
by
AlessandraPascale
- opened
README.md
CHANGED
@@ -14,7 +14,7 @@ H2{color:DarkOrange !important;}
|
|
14 |
p{color:Black !important;}
|
15 |
</style>
|
16 |
|
17 |
-
# Wikipedia
|
18 |
|
19 |
<!-- Provide a quick summary of the dataset. -->
|
20 |
|
@@ -25,7 +25,7 @@ p{color:Black !important;}
|
|
25 |
|
26 |
|
27 |
|
28 |
-
Wikipedia
|
29 |
|
30 |
This dataset card has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1).
|
31 |
|
@@ -35,7 +35,7 @@ This dataset card has been generated using [this raw template](https://github.co
|
|
35 |
|
36 |
<!-- Provide a longer summary of what this dataset is. -->
|
37 |
|
38 |
-
Wikipedia
|
39 |
|
40 |
Each instance consists of a question, a pair of contradictory passages extracted from Wikipedia, and two distinct answers, each derived from on the passages. The pair is annotated by a human annotator who identify where the conflicted information is and what type of conflict is observed. The annotator then produces a set of questions related to the passages with different answers reflecting the conflicting source of knowledge.
|
41 |
|
@@ -91,7 +91,7 @@ N/A.
|
|
91 |
|
92 |
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
|
93 |
|
94 |
-
Wikipedia
|
95 |
The description of each key (when the instance contains two questions) is as follows:
|
96 |
|
97 |
- **title:** Title of article.
|
|
|
14 |
p{color:Black !important;}
|
15 |
</style>
|
16 |
|
17 |
+
# Wikipedia contradict benchmark
|
18 |
|
19 |
<!-- Provide a quick summary of the dataset. -->
|
20 |
|
|
|
25 |
|
26 |
|
27 |
|
28 |
+
Wikipedia contradict benchmark is a dataset consisting of 253 high-quality, human-annotated instances designed to assess LLM performance when augmented with retrieved passages containing real-world knowledge conflicts. The dataset was created intentionally with that task in mind, focusing on a benchmark consisting of high-quality, human-annotated instances.
|
29 |
|
30 |
This dataset card has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/datasetcard_template.md?plain=1).
|
31 |
|
|
|
35 |
|
36 |
<!-- Provide a longer summary of what this dataset is. -->
|
37 |
|
38 |
+
Wikipedia contradict benchmark is a QA-based benchmark consisting of 253 human-annotated instances that cover different types of real-world knowledge conflicts.
|
39 |
|
40 |
Each instance consists of a question, a pair of contradictory passages extracted from Wikipedia, and two distinct answers, each derived from on the passages. The pair is annotated by a human annotator who identify where the conflicted information is and what type of conflict is observed. The annotator then produces a set of questions related to the passages with different answers reflecting the conflicting source of knowledge.
|
41 |
|
|
|
91 |
|
92 |
<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->
|
93 |
|
94 |
+
Wikipedia contradict benchmark is given in JSON format to store the corresponding information, so researchers can easily use our data. There are 253 instances in total.
|
95 |
The description of each key (when the instance contains two questions) is as follows:
|
96 |
|
97 |
- **title:** Title of article.
|