Datasets:

Languages:
English
ArXiv:
Tags:
License:
question1
stringlengths
1
627
question2
stringlengths
1
1.18k
label
class label
2 classes
idx
int32
0
364k
How is the life of a math student? Could you describe your own experiences?
Which level of prepration is enough for the exam jlpt5?
0not_duplicate
0
How do I control my horny emotions?
How do you control your horniness?
1duplicate
1
What causes stool color to change to yellow?
What can cause stool to come out as little balls?
0not_duplicate
2
Where can I find a power outlet for my laptop at Melbourne Airport?
Would a second airport in Sydney, Australia be needed if a high-speed rail link was created between Melbourne and Sydney?
0not_duplicate
4
How not to feel guilty since I am Muslim and I'm conscious we won't have sex together?
I don't beleive I am bulimic, but I force throw up atleast once a day after I eat something and feel guilty. Should I tell somebody, and if so who?
0not_duplicate
5
How is air traffic controlled?
How do you become an air traffic controller?
0not_duplicate
6
What is the best self help book you have read? Why? How did it change your life?
What are the top self help books I should read?
1duplicate
7
Can I enter University of Melbourne if I couldn't achieve the guaranteed marks in Trinity College Foundation?
University of the Philippines: If I take a second BFA in the UP College of Fine Arts, can I be exempted from gen. ed. or core subjects?
0not_duplicate
8
Do you need a passport to go to Jamaica from the United States?
How can I move to Jamaica?
0not_duplicate
9
What is the district of Edgware and how does the lifestyle compare to the London Borough of Islington?
What is the county of Edgware and how does the lifestyle compare to the London Borough of Enfield?
0not_duplicate
10
What will be Hillary Clinton's policy towards India if she becomes president?
What will be Hilary Clinton's policy towards India if she become President?
1duplicate
11
What is the responsibility of SAP ERP key user?
What is a qualified SAP ERP key user?
0not_duplicate
12
Which is the best book to study TENSOR for general relativity from basic?
Which is the best book for tensor calculus?
1duplicate
13
How is being gay or lesbian less moral than divorce?
"Why do a lot of theists and agnostics confuse mainstream atheistic thought with ""positive atheism""?"
0not_duplicate
14
How do you thank a Disneyland cast member?
How can I go to Disneyland with little money?
0not_duplicate
15
What are the coolest Android hacks and tricks you know?
What are some cool hacks for Android phones?
1duplicate
16
Which are the best motivational videos?
What are some of the best motivational clips?
1duplicate
18
How do I lose weight fast?
What is the best way to reduce weight fast?
1duplicate
19
If a die is rolled, what is the probability that the number is greater than 4?
If a die is rolled. what is the probability that the number on top is a 3?
0not_duplicate
20
Whether alloy are only isotropic and homogeneous like metals, or alloys also exhibit orthotropic/anisotropic and heterogeneous like CompositeMaterials?
What is the best backend for my app?
0not_duplicate
22
How can I make me believe that everything is going good in life and get satisfaction when nothing is going right?
What type of government does France currently have and how has it benefited the country?
0not_duplicate
23
How does an IQ test work and what is determined from an IQ test?
How does IQ test works?
1duplicate
24
Is it safe to use Xiaomi mobile phones?
Is it safe or unsafe to use Xiaomi Products?
1duplicate
25
Fetch jobs from job portals through API calls?
What are some creative ideas for arranging a freshers' party?
0not_duplicate
26
What are the best books on cosmology?
Which is the best book for cosmology?
1duplicate
27
Why did it take so long for NASA to find water on Mars?
How much chances are there that NASA already knew that there is water on Mars?
1duplicate
28
My Galaxy ace is hang?
Why are the people on Staten Island are racist?
0not_duplicate
29
Where can I learn to invest in stocks?
How can I learn more about stocks?
1duplicate
30
What are major and defining characteristics of Americans?
What are some good characteristics of the American culture?
0not_duplicate
31
What's the best way to spend a long weekend?
How do I spend my long weekend in an effective way?
1duplicate
32
How do people join ISIS?
Why do people join ISIS?
1duplicate
34
What are some of the most beautiful houses in the world?
Which are some of the most beautiful houses around the world?
1duplicate
36
What are the options if my daughter not got seat in IIT or NIT. At present she is under coaching in Allen, Kota. This year she is in class 11th.?
Who is won indutal medal?
0not_duplicate
37
What is the best lesson we should learn from life?
What's the most important lesson about life?
1duplicate
38
Who are the best technology recruiters in San Diego and Orange County?
Which are the best recruiters for technology executives in the san diego Area?
0not_duplicate
39
What is the best answer to what's up?
What is the best answer for 'Hmmm'?
0not_duplicate
40
If Trump were elected, would he pardon Edward Snowden?
What is Trump's take on Edward Snowden?
1duplicate
41
What would be the scenario of COE placement in Thapar University by 2020 considering the no. of students from CAG and CML too getting pushed in COE?
I got selected in Infosys via campus placement in September 2015 and received my letter of intent in June 2016. When can I expect the offer letter?
0not_duplicate
42
Is there such a thing as a political center? How does it look like and can it be achieved?
There is a good looking guy that acts like he is by boyfriend and that we have a thing. What does it mean?
0not_duplicate
44
What would happen if water disappeared from the earth?
What would happen if ants disappeared from the Earth?
0not_duplicate
45
What was your weirdest childhood interest?
Who are some artists with interesting or inspiring childhoods?
0not_duplicate
46
What's the easiest way to learn Java programs?
What is the easiest way to learn java programming?
1duplicate
47
"Is ""rightness"" just a judgment of an observer or is it ever a property of an action?"
Does the end justify the means?
0not_duplicate
48
Bihar assembly election 2015: why the NDA- led government was not successful in their approach. In future what things should they keep in mind so that they don't make the same mistake again?
Will Narendra Modi win 400+ seat in L S 2019 ?
0not_duplicate
49
What are the best websites for entrepreneurs?
What are the best websites for entrepreneur?
1duplicate
50
What is the National nanotechnology initiative?
What is the lead time for SSN4EGS411 board?
0not_duplicate
51
What are overlap lengths for columns, beam, slab etc?
What is the length of rebars on beams between slabs?
0not_duplicate
53
What type of government did Sparta have?
What types of government did Aristotle want?
0not_duplicate
54
Is it safe to travel to Italy now?
Is it safe to travel through Italy?
1duplicate
56
Is Quora a parallel universe?
Are there any parallel universes?
0not_duplicate
57
Why is circuit switching important?
What is the difference between a virtual circuit and a circuit switch?
0not_duplicate
58
What are some misconceptions, and truths that are believed about Iran?
What are some things people believe about Ireland but are laughably far from the truth?
0not_duplicate
59
Why do most Bollywood movies contain too many sex scenes? Is it because the Bollywood audience are so fond of sex? Are they always horny?
What are the different types of nuclear families? How do they all differ?
0not_duplicate
60
How should I make myself brave?
How can I be more brave?
1duplicate
61
Which is the best Nuru massage parlour in Bangkok?
What is Nuru massage?
0not_duplicate
62
What are the reasons that people dislike Hillary Clinton?
Why do so many people say Hillary Clinton is evil?
1duplicate
63
How do obtain telegram groups link?
I created a Telegram group but I could not find the option to Add Admins. Why?
0not_duplicate
64
Why does the Malayalam movie Premam create such a buzz?
Do Tamils usually watch Malayalam movies?
0not_duplicate
67
What the best science documentaries?
What are some good science documentaries?
1duplicate
70
Why didn't intelligence develop in its highest form in underwater creatures?
How did most creatures develop noses? What was the starting point?
0not_duplicate
71
Why do we, as human beings, use water for?
What do we use water for?
1duplicate
72
How many Parrotheads are there in USA and Canada?
Which technology will win the OLED vs LCD battle?
0not_duplicate
74
Why are nice people always left alone?
How do I accept that I will always be alone?
0not_duplicate
75
What are some examples of parasitic bacteria?
What are some examples of typical bacteria?
0not_duplicate
77
Do employees at Select Income REIT have a good work-life balance? Does this differ across positions and departments?
Do employees at Pennsylvania REIT have a good work-life balance? Does this differ across positions and departments?
0not_duplicate
79
What if every human can read the other person's thoughts? How would life be in such a scenario?
What would happen if every human being can read the thoughts of all other human beings?
1duplicate
80
Why do my dreams are not coming true?
Can dreams come true?
0not_duplicate
82
How will the ban on existing 500 and 1000 rupee note affect India? What are the pros and cons?
What are the advantages and disadvantages of 500 and 1000 rupees ban in India?
1duplicate
83
What are the life lessons that Batman teaches us?
What are the life lessons one can learn from Batman?
1duplicate
85
What are some neurogaming startups?
What are startups?
0not_duplicate
86
What is the remainder when [math]x^{1999}[/math] is divided by [math](x^2-1)[/math] ?
A polynomial leaves remainder [math]2[/math] when divided by [math]x-1[/math] and remainder [math]1[/math] when divided by [math]x-2[/math]. If the polynomial is divided by [math](x-1)(x-2)[/math], then what would be the remainder?
0not_duplicate
87
Can mechanical energy be conserved?
When and how is mechanical energy not conserved?
1duplicate
88
How can I lose 4kg weight?
What are the ways of losing weight?
1duplicate
91
Which is the most independent country?
What country has the most attractive women -- either in absolute terms or in density?
0not_duplicate
92
What are some reviews of www.cosmicpetrochem.com?
What is your review of www.buttermyresume.com?
0not_duplicate
93
How do I create a stage name for myself?
How can I create a stage name?
1duplicate
95
? to be deleted
Did Hitler underestimate the jewish mafia-nation?
0not_duplicate
96
Can you grow a tree in zero gravity?
How do trees grow in a zero gravity environment?
1duplicate
98
What is your review of South Indian Queens?
What is your review of South Indians?
0not_duplicate
100
"Why does Quora mark my perfectly semantic question as, ""Needs Improvement""?"
"What should I do if Quora marks my question as ""Needs Improvement""?"
1duplicate
101
Why can't I play high quality games like N.O.V.A 3 in bluestacks?
If I liked Skyrim, what other games would I like?
0not_duplicate
102
How can I meet new people on Skype?
What is the best way to meet new people?
0not_duplicate
103
What is it like to live in Forest Hill, San Francisco?
What is it like to live in the Chinatown or Nob Hill areas of San Francisco?
0not_duplicate
104
What do you think about Barack Obama?
What is Barack Obama doing now?
0not_duplicate
105
What is the business model for intent.com?
What is the business model for wooroll.com?
0not_duplicate
107
Is there any bar in Dublin where one can meet mature ladies who like young boys?
What is the best way to meet women?
0not_duplicate
110
How will Donald trump's Victory would affect India's relationship with USA?
What is the impact of Donald Trump's victory on Indian interests?
1duplicate
111
What is a Gauchais Reaction?
What is the first reaction of a girl when a boy proposes?
0not_duplicate
112
What's the difference between a promise and a callback in Javascript?
What is the difference between Javascript, JSP, Node.js and EJS? From where should I begin?
0not_duplicate
113
Why is pewter so expensive, and how does its properties compare to those of gold?
Why is pewter so expensive, and how does its properties compare to those of aluminum?
0not_duplicate
114
What is the in hand salary of SBI so?
What will be my in Hand salary if I have?
0not_duplicate
116
How can I improve my English in all aspects?
How can I continue to improve my English?
1duplicate
117
How do I prepare for interviews?
How can I prepare for my interview?
1duplicate
118
How will Donald Trump's presidency affect international students?
What would be the effect of Trump's presidency on international Master's students who plan to work in the US after completing the degree?
1duplicate
120
What are stars made of?
What are stars made out of?
1duplicate
121
What companies are similar to IDEO in Paris?
What companies are similar to IDEO?
0not_duplicate
122
Why do colleges change to university?
Which are the most modern courses in engineering?
0not_duplicate
123
How can I lose 10 Kilos?
How should I lose weight?
1duplicate
124
What facts about the United States do foreigners not believe until they come to America?
How do foreigners who have settled in India feel about India?
0not_duplicate
126
What's the layman explanation of Godel's incompleteness theorem?
What is an explanation of Gödel's Incompleteness Theorems suitable for a 10 year-old?
0not_duplicate
128

Dataset Card for Bias-amplified Splits for QQP

Dataset Summary

Bias-amplified splits is a novel evaluation framework to assess model robustness, by amplifying dataset biases in the training data and challenging models to generalize beyond them. This framework is defined by a bias-amplified training set and a hard, anti-biased test set, which we automatically extract from existing datasets using model-based methods.

Our experiments show that the identified anti-biased examples are naturally challenging for models, and moreover, models trained on bias-amplified data exhibit dramatic performance drops on anti-biased examples, which are not mitigated by common approaches to improve generalization.

Here we apply our framework to the Quora Question Pairs dataset (QQP), a dataset composed of question pairs where the task is to determine if the questions are paraphrases of each other (have the same meaning).

Our evaluation framework can be applied to any existing dataset, even those considered obsolete, to test model robustness. We hope our work will guide the development of robust models that do not rely on superficial biases and correlations.

Evaluation Results (DeBERTa-large)

For splits based on minority examples:
Training Data \ Test Data Original test Anti-biased test
Original training split 93.0 77.6
Biased training split 87.0 36.8
For splits based on partial-input model:
Training Data \ Test Data Original test Anti-biased test
Original training split 93.0 81.3
Biased training split 90.3 63.9

Loading the Data

from datasets import load_dataset

# choose which bias detection method to use for the bias-amplified splits: either "minority_examples" or "partial_input"
dataset = load_dataset("bias-amplified-splits/qqp", "minority_examples")

# use the biased training split and anti-biased test split
train_dataset = dataset['train.biased']
eval_dataset = dataset['validation.anti_biased']

Dataset Structure

Data Instances

Data instances are taken directly from QQP (GLUE version), and re-split into biased and anti-biased subsets. Here is an example of an instance from the dataset:

{
  "idx": 56,
  "question1": "How do I buy used car in India?",
  "question2": "Which used car should I buy in India?",
  "label": 0
}

Data Fields

  • idx: unique identifier for the example within its original data splits (e.g., validation set)
  • question1: a question asked on Quora
  • question2: a question asked on Quora
  • label: one of 0 and 1 (not duplicate and duplicate)

Data Splits

Bias-amplified splits require a method to detect biased and anti-biased examples in datasets. We release bias-amplified splits based created with each of these two methods:

  • Minority examples: A novel method we introduce that leverages representation learning and clustering for identifying anti-biased minority examples (Tu et al., 2020)—examples that defy common statistical patterns found in the rest of the dataset.
  • Partial-input baselines: A common method for identifying biased examples containing annotation artifacts in a dataset, which examines the performance of models that are restricted to using only part of the input. Such models, if successful, are bound to rely on unintended or spurious patterns in the dataset.

Using each of the two methods, we split each of the original train and test splits into biased and anti-biased subsets. See the paper for more details.

Minority Examples

Dataset Split Number of Instances in Split
Train - biased 297735
Train - anti-biased 66111
Validation - biased 32968
Validation - anti-biased 7462

Partial-input Baselines

Dataset Split Number of Instances in Split
Train - biased 297735
Train - anti-biased 66111
Validation - biased 33084
Validation - anti-biased 7346

Dataset Creation

Curation Rationale

NLP models often rely on superficial cues known as dataset biases to achieve impressive performance, and can fail on examples where these biases do not hold. To develop more robust, unbiased models, recent work aims to filter bisased examples from training sets. We argue that in order to encourage the development of robust models, we should in fact amplify biases in the training sets, while adopting the challenge set approach and making test sets anti-biased. To implement our approach, we introduce a simple framework that can be applied automatically to any existing dataset to use it for testing model robustness.

Annotations

Annotation process

No new annotations are required to create bias-amplified splits. Existing data instances are split into biased and anti-biased splits based on automatic model-based methods to detect such examples.

Considerations for Using the Data

Social Impact of Dataset

Bias-amplified splits were created to promote the development of robust NLP models that do not rely on superficial biases and correlations, and provide more challenging evaluation of existing systems.

Discussion of Biases

We propose to use bias-amplified splits to complement benchmarks with challenging evaluation settings that test model robustness, in addition to the dataset’s main training and test sets. As such, while existing dataset biases are amplified during training with bias-amplified splits, these splits are intended primarily for model evaluation, to expose the bias-exploiting behaviors of models and to identify more robsut models and effective robustness interventions.

Additional Information

Dataset Curators

Bias-amplified splits were introduced by Yuval Reif and Roy Schwartz from the Hebrew University of Jerusalem.

QQP data was released by Quora and released under the GLUE benchmark.

Citation Information

@misc{reif2023fighting,
    title = "Fighting Bias with Bias: Promoting Model Robustness by Amplifying Dataset Biases",
    author = "Yuval Reif and Roy Schwartz",
    month = may,
    year = "2023",
    url = "https://arxiv.org/pdf/2305.18917",
}

Source dataset:

@inproceedings{wang2019glue,
  title={{GLUE}: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding},
  author={Wang, Alex and Singh, Amanpreet and Michael, Julian and Hill, Felix and Levy, Omer and Bowman, Samuel R.},
  note={In the Proceedings of ICLR.},
  year={2019}
}
Downloads last month
2
Edit dataset card