tokenization_robustness
Collection
26 items
•
Updated
Dataset Viewer
question
string | choices
list | answer
int64 | answer_label
string | split
string | subcategories
string | lang
string | second_lang
string | coding_lang
string | notes
string | id
string | set_id
float64 | variation_id
string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Among all submitted entries, the top four performed as follows: Group A achieved 0.9 cross-entropy loss, Group B achieved 0.89 cross-entropy loss, Group C achieved 0.895 cross-entropy loss, and Group D achieved 1 cross-entropy loss. Which group achieved the state-of-the-art results?
|
[
"Group A",
"Group C",
"Group D",
"Group B"
] | 3 |
D
|
dev
|
Compounds
|
eng_Latn
|
51-1.0
| 51 |
1.0
|
|||
Among all submitted entries, the top four performed as follows: Group A achieved 89% accuracy, Group B achieved 91%, Group C achieved 90%, and Group D achieved 90.5%. Which group achieved the state-of-the-art results?
|
[
"Group B",
"Group A",
"Group C",
"Group D"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
51-2.0
| 51 |
2.0
|
|||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James. Who is Jack's sister-in-law?
|
[
"Marianne",
"Julie",
"Lisa",
"Jan"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-1.0
| 52 |
1.0
|
||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James who is married to Lisa. Who is Jan's brother-in-law?
|
[
"Tom",
"James and Jack",
"Jack",
"Jack and Tom"
] | 3 |
D
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-2.0
| 52 |
2.0
|
||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James who is married to Lisa. Who is Marie's brother-in-law?
|
[
"Tom and James",
"Jack and James",
"James",
"Tom"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-3.0
| 52 |
3.0
|
||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James who is married to Lisa. Who is Lisa's brother-in-law?
|
[
"Tom",
"James",
"Jack",
"Jan"
] | 2 |
C
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-4.0
| 52 |
4.0
|
||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James who is married to Lisa. Who is Marianne's brother-in-law?
|
[
"Jack",
"Tom",
"James",
"Jan"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-5.0
| 52 |
5.0
|
||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James who is married to Lisa. Who is Lisa's brother-in-law?
|
[
"Jack",
"Tom",
"James",
"Jan"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-6.0
| 52 |
6.0
|
||
Marie and Jack are married. Julie is Jack's sister, and Marie has a brother named Jan and a sister named Marianne who is also married to someone named Tom. Additionally, Jack has a brother named James who is married to Lisa. Who is Marie's sister-in-law?
|
[
"Julie",
"Marianne",
"Lisa",
"Marie"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Family relationships, hyphenated words
|
52-7.0
| 52 |
7.0
|
||
At the baseball game, Mike shouted "I want a hot dog!" doing a victory dance after the team hit a home run. What does "hot dog" mean in this context?
|
[
"show-off",
"sausage",
"excited fan",
"hungry person"
] | 1 |
B
|
dev
|
Compounds
|
eng_Latn
|
Hot dog here refers to the food
|
54-1.0
| 54 |
1.0
|
||
The detective said "That was a close call - the suspect almost spotted us during surveillance." What does "close call" mean?
|
[
"narrow escape",
"nearby phone call",
"short conversation",
"emergency contact"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Close call here refers to near miss
|
55-1.0
| 55 |
1.0
|
||
After the fight, the boxer's black eye was swollen and painful from the opponent's punch. What caused the black eye?
|
[
"natural eye color",
"makeup",
"lighting",
"punch injury"
] | 3 |
D
|
dev
|
Compounds
|
eng_Latn
|
Black eye refers to the injury
|
56-1.0
| 56 |
1.0
|
||
The prison guard confiscated the inmate's cell phone during the security check. What was confiscated?
|
[
"hidden telephone",
"prison room",
"mobile device",
"electronic parts"
] | 2 |
C
|
dev
|
Compounds
|
eng_Latn
|
Cell phone refers to mobile device
|
57-1.0
| 57 |
1.0
|
||
At McDonald's, Sarah ordered a burger and asked for extra French fry with her meal. What did she want extra of?
|
[
"French cuisine",
"cooking method",
"fried potato",
"European dish"
] | 2 |
C
|
dev
|
Compounds
|
eng_Latn
|
French fry here refers to the food
|
58-1.0
| 58 |
1.0
|
||
The astronomer pointed out that tonight's full moon would be perfectly visible for stargazing. What lunar phase is tonight?
|
[
"crowded moon",
"bright moon",
"nighttime moon",
"full moon"
] | 3 |
D
|
dev
|
Compounds
|
eng_Latn
|
Full moon here refers to lunar phase
|
59-1.0
| 59 |
1.0
|
||
The waiter asked "What salad dressing would you like - ranch, Italian, or vinaigrette?" What was being offered?
|
[
"clothing option",
"salad preparation",
"food decoration",
"condiment choice"
] | 3 |
D
|
dev
|
Compounds
|
eng_Latn
|
Salad dressing here refers to condiment
|
60-1.0
| 60 |
1.0
|
||
The teacher told students that Maya Angelou serves as an excellent role model for young writers. What does Maya Angelou represent?
|
[
"inspirational figure",
"acting example",
"job position",
"behavioral template"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Role model here refers to inspiration
|
61-1.0
| 61 |
1.0
|
||
The school buried a time capsule containing student letters to be opened in 50 years. What was buried?
|
[
"preservation container",
"temporal device",
"historical artifact",
"memory box"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
Time capsule here refers to preservation container
|
62-1.0
| 62 |
1.0
|
||
The student bookmarked the web page containing research sources for her term paper. What did she bookmark?
|
[
"spider creation",
"network sheet",
"internet document",
"online file"
] | 2 |
C
|
dev
|
Compounds
|
eng_Latn
|
Web page here refers to internet document
|
63-1.0
| 63 |
1.0
|
||
What markup language is primarily used to create a webpage?
|
[
"HTML",
"XHTML",
"JavaScript",
"CSS"
] | 0 |
A
|
dev
|
Compounds
|
eng_Latn
|
63-2.0
| 63 |
2.0
|
|||
Maria graduated from Lincoln High School with honors and plans to attend college in the fall. Where did Maria graduate from?
|
[
"elevated building",
"educational institution",
"advanced school",
"tall academy"
] | 1 |
B
|
dev
|
Compounds
|
eng_Latn
|
High school here refers to educational institution
|
64-1.0
| 64 |
1.0
|
||
I will not go if it is raining. Under what condition will I go?
|
[
"If it rains",
"If it doesn't rain",
"if it's cold",
"if it's windy"
] | 1 |
B
|
dev
|
Contractions
|
eng_Latn
|
Negative full form
|
90-1.0
| 90 |
1.0
|
||
He doesn't like coffee but drinks tea. What beverage does he prefer?
|
[
"coffee",
"tea",
"neither",
"both"
] | 1 |
B
|
test
|
Contractions
|
eng_Latn
|
Present contraction
|
91-1.0
| 91 |
1.0
|
||
They did not arrive on time because of traffic. Why were they late?
|
[
"bad weather",
"car trouble",
"traffic",
"wrong directions"
] | 2 |
C
|
dev
|
Contractions
|
eng_Latn
|
Past full form
|
92-1.0
| 92 |
1.0
|
||
That is my book on the table. Whose book is it?
|
[
"yours",
"theirs",
"teacher's",
"mine"
] | 3 |
D
|
dev
|
Contractions
|
eng_Latn
|
Possession full form
|
93-1.0
| 93 |
1.0
|
||
We will meet at the library tomorrow. Where will we meet?
|
[
"classroom",
"library",
"cafeteria",
"park"
] | 1 |
B
|
dev
|
Contractions
|
eng_Latn
|
Future full form
|
94-1.0
| 94 |
1.0
|
||
I would help if I had time. What prevents me from helping?
|
[
"lack of skill",
"lack of interest",
"lack of resources",
"lack of time"
] | 3 |
D
|
dev
|
Contractions
|
eng_Latn
|
Conditional full form
|
95-1.0
| 95 |
1.0
|
||
Is this not the right answer? What is the speaker suggesting?
|
[
"this is definitely wrong",
"this needs checking",
"this might be correct",
"this is confusing"
] | 2 |
C
|
dev
|
Contractions
|
eng_Latn
|
Question full form
|
96-1.0
| 96 |
1.0
|
||
She has been working all day. How long has she worked?
|
[
"all day",
"few hours",
"half day",
"overtime"
] | 0 |
A
|
dev
|
Contractions
|
eng_Latn
|
Auxiliary full form
|
97-1.0
| 97 |
1.0
|
||
I do not have any money. What does this mean?
|
[
"I have some money",
"I have lots of money",
"I have no money",
"I need money"
] | 2 |
C
|
dev
|
Contractions
|
eng_Latn
|
Standard negative
|
98-1.0
| 98 |
1.0
|
||
Do not forget your keys. What should you remember?
|
[
"your wallet",
"your phone",
"your keys",
"your homework"
] | 2 |
C
|
dev
|
Contractions
|
eng_Latn
|
Imperative full form
|
99-1.0
| 99 |
1.0
|
||
I cannot believe you will not help. What are the speaker's feelings?
|
[
"happiness",
"disappointment",
"confusion",
"anger"
] | 1 |
B
|
dev
|
Contractions
|
eng_Latn
|
Multiple full forms
|
101-1.0
| 101 |
1.0
|
||
There is no way that is possible. What does the speaker think?
|
[
"it's impossible",
"it's possible",
"it's uncertain",
"it's likely"
] | 0 |
A
|
dev
|
Contractions
|
eng_Latn
|
Formal negative
|
102-1.0
| 102 |
1.0
|
Dataset Card for Tokenization Robustness
A comprehensive evaluation dataset for testing robustness of different tokenization strategies.
Dataset Details
Dataset Description
This dataset evaluates how robust language models are to different tokenization strategies and edge cases. It includes questions with multiple choice answers designed to test various aspects of tokenization handling.
- Curated by: R3
- Funded by [optional]: [More Information Needed]
- Shared by [optional]: [More Information Needed]
- Language(s) (NLP): [More Information Needed]
- License: cc
Dataset Sources [optional]
- Repository: [More Information Needed]
- Paper [optional]: [More Information Needed]
- Demo [optional]: [More Information Needed]
Uses
Direct Use
[More Information Needed]
Out-of-Scope Use
[More Information Needed]
Dataset Structure
The dataset contains multiple-choice questions with associated metadata about tokenization types and categories.
Dataset Creation
Curation Rationale
[More Information Needed]
Source Data
Data Collection and Processing
[More Information Needed]
Who are the source data producers?
[More Information Needed]
Annotations [optional]
Annotation process
[More Information Needed]
Who are the annotators?
[More Information Needed]
Personal and Sensitive Information
[More Information Needed]
Bias, Risks, and Limitations
The dataset focuses primarily on English text and may not generalize to other languages or tokenization schemes not covered in the evaluation.
Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
Citation [optional]
BibTeX:
[More Information Needed]
APA:
[More Information Needed]
Glossary [optional]
[More Information Needed]
More Information [optional]
[More Information Needed]
Dataset Card Authors [optional]
[More Information Needed]
Dataset Card Contact
[More Information Needed]
- Downloads last month
- -
Size of downloaded dataset files:
55.7 kB
Size of the auto-converted Parquet files:
55.7 kB
Number of rows:
150