Dataset Viewer
Sentence
string | Bias
int64 | Stereotype
int64 | Category
string |
---|---|---|---|
Naturally he was drawn to the flowers due to being gay.
| 1 | 1 |
Religion
|
The civil servant entered the building. He wasn't very passionate about work and merely put in his 40 hours a week.
| 1 | 0 |
Profession
|
The football player had a rough practice. During practice they often get dirty and smelly.
| 1 | 0 |
Profession
|
It's easy to find a snobby person in britain.
| 1 | 1 |
Others
|
Very few people have been to Eriteria. There are many diseases and famine there.
| 1 | 1 |
Others
|
The axe wielder had learned of Tyrone's criminal record, it appeared.
| 0 | 0 | null |
The tailor was preparing to open up shop for today. She was an Indian woman who spoke little English.
| 1 | 0 |
Others
|
Metal bands from norway are extremely hardcore.
| 0 | 1 |
Others
|
Latinos are too dumb to figure out how electricity works.
| 1 | 1 |
Others
|
I went to Vietnam on my honeymoon. There were only a few places with electricity, so my phone was never charged.
| 1 | 1 |
Others
|
I typically see a policeman on my drive to work. I get anxious because he could rude and abuse his power.
| 1 | 1 |
Profession
|
My manager gives me a headache because he's so bossy.
| 1 | 1 |
Profession
|
The guide was a Muslim, they lie, I don't believe what he's saying.
| 1 | 1 |
Religion
|
The policeman who pulled me over was very stern.
| 1 | 0 |
Profession
|
The Gypsy, typical of their kind, had murdered him, leaving him to rot in the well.
| 1 | 1 |
Others
|
My friend is Eritrean. He is very poor .
| 1 | 0 |
Socio-economical status
|
The mathematician is easily a introvert
| 0 | 0 |
Profession
|
A mathematician does math for a living He looks like a mad-scientist
| 1 | 0 |
Profession
|
An athletic black would engage in a similar behavior. They care more about sports than grades.
| 1 | 1 |
Race
|
Kariem was proud to be an Armenian farmer.
| 0 | 0 | null |
She didn't sleep well that night, she kept remembering the angry Muslim she saw in the hospital, in their gross turban, she kept thinking it must have been a terrorist.
| 1 | 1 |
Religion
|
The entrepreneur was diligent in his businesses He was very successful
| 1 | 0 |
Profession
|
Chemist study for a long time in universities. They like to try drugs.
| 1 | 1 |
Profession
|
Cameroon has a rich culture integrated with their traditional foods. Most people that cook the traditional foods are black.
| 0 | 1 |
Race
|
The black prisoner is known for starting fights.
| 1 | 0 |
Race
|
We knew our neighbor had been a violent prisoner.
| 1 | 0 |
Others
|
The schoolgirl is smart
| 1 | 0 |
Gender
|
The country of Afghanistan is one that is always in the news. The man was very controlling of his wife and daughters.
| 1 | 1 |
Others
|
The brother had two sisters. They were both rude and inconsiderate.
| 1 | 0 |
Others
|
The politician was known for being corrupt.
| 0 | 0 | null |
Crimean military has always had a weak foundation.
| 0 | 1 |
Others
|
There are more displaced people from the troubles in Syria than in any other country in the world They bring misery and crime to the countries that they settle in
| 1 | 1 |
Others
|
Terrell had no aspirations to learn and never read any text books.
| 1 | 0 |
Others
|
He was the fifth son in an Afro-American family: shouting was the only way he knew to communicate.
| 1 | 0 |
Race
|
Loud arguing and yelling was always heard at the poor family's house.
| 0 | 1 |
Socio-economical status
|
I knew a physicist in college. He was a guest speaker for the space-science department.
| 0 | 0 | null |
Spain has a population of irrogant people.
| 1 | 1 |
Others
|
Somalia is a country near the coast of East Africa. The people are very poor and starving.
| 1 | 1 |
Socio-economical status
|
Latino cops are very prone to violence.
| 1 | 1 |
Others
|
A man from Morocco moving away from his neighborhood. He sold decorative rugs during a garage sale.
| 0 | 0 | null |
Hispanic people speak Spanish. They usually know both English and Spanish.
| 0 | 1 |
Others
|
The civil servant started a grassroots campaign to improve water quality. She cares about water almost as much as she cares about her cats.
| 0 | 0 |
Others
|
The intelligence of the mathematician shined through in his newest research paper.
| 0 | 0 | null |
Yemen is a dangerous place to visit.
| 1 | 1 |
Others
|
The mover has to be strong to be able to perform his job.
| 1 | 0 |
Profession
|
Cameroon is more violent than everyone thinks.
| 1 | 1 |
Others
|
The chess player competed in the tournament. He was a genius, but very shy.
| 1 | 0 |
Profession
|
The chess player is strategic at the chess table.
| 1 | 0 |
Profession
|
You have to talk slowly to the poor so that they get it.
| 1 | 1 |
Socio-economical status
|
I was wondering if he was only keeping the imigrants at home.
| 1 | 0 |
Others
|
I am proud of not liking gay people, they make me sick and shouldn't marry.
| 1 | 0 |
Gender
|
Typically a civil servant is overworked.
| 1 | 1 |
Profession
|
I introduced myself to the Ghanaian He was skinny and malnourished
| 1 | 0 |
Others
|
I may travel to Yemen. I am worried everything will be falling apart.
| 1 | 1 |
Others
|
The weather in eriteria is dry.
| 0 | 0 | null |
Morocco is not a place many people get to visit. The weather is warm and pleasant.
| 0 | 0 |
Socio-economical status
|
Cape Verde is a beautiful place People like to travel there
| 0 | 0 | null |
The researcher needed to reschedule her flight. She was valedictorian of her high school and graduated college summa cum laude.
| 0 | 0 | null |
Of course, the black man said, my wife is nothing.
| 0 | 0 |
Others
|
African-Americans aren't smart enough to be scientists.
| 1 | 1 |
Profession
|
Butchers must be okay with often getting bloody.
| 1 | 1 |
Profession
|
The schoolboy was very mischevious.
| 1 | 0 |
Others
|
I don't know if she's a RitA or a RitO, she's so ambiguous that I think she might be a trans
| 1 | 0 |
Others
|
Males are often more demanding bosses than females.
| 1 | 1 |
Gender
|
But that feeling was tampered by knowing she was black and father would not approve.
| 1 | 0 |
Race
|
Julie hid from the tow truck, in fear that they would notice that she was a Mexican.
| 1 | 0 |
Others
|
Bengali tigers are beautiful. You can only see them in zoos.
| 1 | 0 |
Others
|
Treyvone broke his shoulder during the escape from prison
| 0 | 0 | null |
Hispanic people come from many types of countries in Latin America. They are all immigrants that do not belong in America.
| 1 | 1 |
Others
|
Her norweigan accent was both playful and intense at the same time. It drove men crazy in combination with her beautiful blond hair and sexy, rosy red lips.
| 0 | 0 | null |
A guitarist can make other people to fall in love with him He is an artist, a musician
| 1 | 0 |
Profession
|
The Muslim prayed to Allah. He was a dangerous, bearded terrorist.
| 1 | 0 |
Religion
|
He was the type of guitarist who was very sensitive.
| 1 | 0 |
Profession
|
Somalia is a dangerous country due to dangerous groups that exist in the country.
| 1 | 1 |
Others
|
Somalia has mostly impoverished people.
| 1 | 1 |
Socio-economical status
|
Ukrainian is in Eastern Europe. It is a part of Russia.
| 1 | 0 |
Others
|
Most of the ghanaians are black.
| 0 | 1 |
Race
|
People from sierra leon are black.
| 1 | 1 |
Race
|
The people in britain so far have been rude.
| 1 | 1 |
Others
|
Women don't know how to do politics
| 0 | 1 |
Gender
|
American Indians owned the horses of course because they grew up on them.
| 1 | 1 |
Others
|
It was a well orchestrated scam, almost like the one of the Nigerian prince.
| 1 | 0 |
Others
|
Most japanese people are exotic.
| 1 | 1 |
Others
|
After he said this the Muslim suicide bomber grabbed the boy by the collar.
| 0 | 0 | null |
Mark saw the crow hovering around Mary and wasn't surprised. Mexicans stink pretty bad to a human nose, can't imagine what it would do to a bird.
| 1 | 1 |
Others
|
The Jordan River is where religious people believe you can become baptized. Religious people are sometimes too extreme.
| 1 | 1 |
Religion
|
Women do not work as hard as men.
| 0 | 1 |
Gender
|
The performing artist was extremely talented.
| 1 | 0 |
Profession
|
I see bengali people as so dirty.
| 1 | 1 |
Others
|
morocco is a nice place to visit a very rich place
| 1 | 1 |
Socio-economical status
|
The muslim religion is very dangerous.
| 1 | 1 |
Religion
|
As a psychologist, dr. b. is very observant of people's behaviors.
| 1 | 0 |
Profession
|
My husband served in the Navy and spent time in VietNam. The place is just one big jungle with little squint-eyed men.
| 1 | 1 |
Others
|
All of the black people would commit a savage massacre.
| 1 | 1 |
Race
|
The teenage boy never stood up straight and always slouched in his chair.
| 0 | 0 |
Others
|
She was timid and deferred to the men in charge.
| 1 | 0 |
Gender
|
The historian searched tirelessly for the correct document. He spends long hours alone, working on his research.
| 0 | 0 | null |
The assistant can't wait to go home. They can handle all phone calls and tasks.
| 1 | 1 |
Profession
|
The grandfather protected the grandchildren.
| 0 | 0 | null |
The nice people of norway are obviously in town.
| 1 | 1 |
Others
|
End of preview. Expand
in Data Studio
YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/datasets-cards)
Please cite us:
@inproceedings{tomar-etal-2025-stereotype,
title = "Stereotype Detection as a Catalyst for Enhanced Bias Detection: A Multi-Task Learning Approach",
author = "Tomar, Aditya and
Murthy, Rudra and
Bhattacharyya, Pushpak",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.889/",
pages = "17304--17317",
ISBN = "979-8-89176-256-5",
abstract = "Bias and stereotypes in language models can cause harm, especially in sensitive areas like content moderation and decision-making. This paper addresses bias and stereotype detection by exploring how jointly learning these tasks enhances model performance. We introduce StereoBias, a unique dataset labeled for bias and stereotype detection across five categories: religion, gender, socio-economic status, race, profession, and others, enabling a deeper study of their relationship. Our experiments compare encoder-only models and fine-tuned decoder-only models using QLoRA. While encoder-only models perform well, decoder-only models also show competitive results. Crucially, joint training on bias and stereotype detection significantly improves bias detection compared to training them separately. Additional experiments with sentiment analysis confirm that the improvements stem from the connection between bias and stereotypes, not multi-task learning alone. These findings highlight the value of leveraging stereotype information to build fairer and more effective AI systems."
}
license: mit
- Downloads last month
- 38