Upload prompt template grounding_accuracy_implicit_span_level.yaml
Browse files
grounding_accuracy_implicit_span_level.yaml
ADDED
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
prompt:
|
2 |
+
template: |-
|
3 |
+
Your task is to check if the Response is accurate to the Evidence.
|
4 |
+
Generate 'Accurate' if the Response is accurate when verified according to the Evidence, or 'Inaccurate' if the Response is inaccurate (contradicts the evidence) or cannot be verified.
|
5 |
+
|
6 |
+
**Query**:
|
7 |
+
|
8 |
+
{{user_request}}
|
9 |
+
|
10 |
+
**End of Query**
|
11 |
+
|
12 |
+
**Evidence**
|
13 |
+
|
14 |
+
{{context_document}}
|
15 |
+
|
16 |
+
**End of Evidence**
|
17 |
+
|
18 |
+
**Response**:
|
19 |
+
|
20 |
+
{{response}}
|
21 |
+
|
22 |
+
**End of Response**
|
23 |
+
|
24 |
+
|
25 |
+
Break down the Response into sentences and classify each one separately, then give the final answer: If even one of the sentences is inaccurate, then the Response is inaccurate.
|
26 |
+
|
27 |
+
For example, your output should be of this format:
|
28 |
+
Sentence 1: <Sentence 1>
|
29 |
+
Sentence 1 label: Accurate/Inaccurate (choose 1)
|
30 |
+
Sentence 2: <Sentence 2>
|
31 |
+
Sentence 2 label: Accurate/Inaccurate (choose 1)
|
32 |
+
Sentence 3: <Sentence 3>
|
33 |
+
Sentence 3 label: Accurate/Inaccurate (choose 1)
|
34 |
+
[...]
|
35 |
+
Final Answer: Accurate/Inaccurate (choose 1)
|
36 |
+
template_variables:
|
37 |
+
- user_request
|
38 |
+
- context_document
|
39 |
+
- response
|
40 |
+
metadata:
|
41 |
+
description: "An evaluation prompt from the paper 'The FACTS Grounding Leaderboard: Benchmarking LLMs’ Ability to Ground
|
42 |
+
Responses to Long-Form Input' by Google DeepMind.\n The prompt was copied from the evaluation_prompts.csv file from
|
43 |
+
Kaggle.\n This specific prompt elicits a binary accurate/non-accurate classifier for the entire response after generating
|
44 |
+
and classifying each sentence separately."
|
45 |
+
evaluation_method: implicit_span_level
|
46 |
+
tags:
|
47 |
+
- fact-checking
|
48 |
+
version: 1.0.0
|
49 |
+
author: Google DeepMind
|
50 |
+
source: https://www.kaggle.com/datasets/deepmind/FACTS-grounding-examples?resource=download&select=evaluation_prompts.csv
|
51 |
+
client_parameters: {}
|
52 |
+
custom_data: {}
|