key
stringlengths 7
86
| messages
listlengths 1
1
| ground_truth
stringlengths 62
341k
| dataset
stringclasses 1
value | constraint_type
stringclasses 1
value | constraint
stringlengths 31
805
|
|---|---|---|---|---|---|
personas_math_easy_5ccsg1396p08a8ovy9vr5189
|
[
{
"content": "Alex lives in the charming neighborhood of German Village in Columbus and enjoys exploring its beautiful parks and historic streets. One day, Alex decides to take a long walk around the area. He starts by walking 3 blocks east to visit a local bakery, then walks 4 blocks south to see the famous Schiller Park. After spending some time at the park, Alex continues walking 2 blocks west to reach a quaint bookstore. Finally, he walks 5 blocks north to return home. How many blocks did Alex walk in total during his outing in German Village? In your response, the letter b should appear at least 18 times.",
"role": "user"
}
] |
[{'instruction_id': ['letters:letter_counting2'], 'kwargs': [{'letter': 'b', 'let_frequency': 18, 'let_relation': 'at least'}]}]
|
ifeval
|
multi
|
In your response, the letter b should appear at least 18 times.
|
ai2-adapt-dev/tulu_v3.9_synthetic_finalresp_wildguardmixtrain_decontaminated_50k_13390
|
[
{
"content": "i want to focus on endurance. Only use unique words in your response, no word should be repeated! Do not include keyword i in the response.",
"role": "user"
}
] |
[{'instruction_id': ['count:count_unique', 'keywords:exclude_word_harder'], 'kwargs': [None, {'keyword': 'i'}]}]
|
ifeval
|
multi
|
Only use unique words in your response, no word should be repeated! Do not include keyword i in the response.
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_78053
|
[
{
"content": "Is it possible to embed an inline search box into a web page which provides similar functionality to the [IE7Pro Inline Search](http://www.ie7pro.com/inline-search.html) or similar plugins for Firefox/Safari? Repeat the phrase Laughter is the best medicine exactly 3 times, transforming it slightly each time by replacing only one word in the center of the phrase.",
"role": "user"
}
] |
[{'instruction_id': ['copy:repeat_phrase'], 'kwargs': [{'phrase': 'Laughter is the best medicine', 'small_n': 3}]}]
|
ifeval
|
multi
|
Repeat the phrase Laughter is the best medicine exactly 3 times, transforming it slightly each time by replacing only one word in the center of the phrase.
|
ai2-adapt-dev/numinamath_tir_math_decontaminated_4522
|
[
{
"content": "Bella bought stamps at the post office. Some of the stamps had a snowflake design, some had a truck design, and some had a rose design. Bella bought 11 snowflake stamps. She bought 9 more truck stamps than snowflake stamps, and 13 fewer rose stamps than truck stamps. How many stamps did Bella buy in all? Start and end your response with the same word (do not write anything after the last word, not even punctuation). There should be 2 paragraphs. Paragraphs are separated with the markdown divider: ***",
"role": "user"
}
] |
[{'instruction_id': ['keywords:start_end', 'paragraphs:paragraphs'], 'kwargs': [None, {}]}]
|
ifeval
|
multi
|
Start and end your response with the same word (do not write anything after the last word, not even punctuation). There should be 2 paragraphs. Paragraphs are separated with the markdown divider: ***
|
personas_math_easy_7ixjrbm37bs0no4cj1ye94gx
|
[
{
"content": "Dr. Quantum, a physicist fascinated by the mysteries of the subatomic world, decides to conduct an experiment involving tiny particles. She has a special container that can hold exactly 200 quantum particles. However, each time she observes the particles, 15% of them behave unpredictably and escape from the container. \n\nDr. Quantum starts her experiment by placing 120 particles in the container. After her first observation, 15% escape. She then adds 20 more particles to the container to continue her research. After her second observation, again, 15% of the particles escape. How many particles are left in the container after these two observations and additions? Answer with one of the following options: ('My answer is yes.', 'My answer is no.', 'My answer is maybe.', 'I think yes.', 'I think no.', 'I think maybe.') Write 3 paragraphs, delimited by the markdown divider: * * *, with exactly 2 sentences each, with exactly 3 words in each sentence. Wrap every word bigram in double angular brackets, such as <<I am>> <<at home>> <<with my>> <<cute dog>>. The first word of each sentence should be the word form. Wrap your entire response with double quotation marks.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:constrained_response', 'count:counting_composition', 'detectable_format:bigram_wrapping', 'first_word:first_word_sent', 'startend:quotation'], 'kwargs': [None, {'n_sent': 2, 'n_words': 3}, None, {'first_word': 'form'}, None]}]
|
ifeval
|
multi
|
Answer with one of the following options: ('My answer is yes.', 'My answer is no.', 'My answer is maybe.', 'I think yes.', 'I think no.', 'I think maybe.') Write 3 paragraphs, delimited by the markdown divider: * * *, with exactly 2 sentences each, with exactly 3 words in each sentence. Wrap every word bigram in double angular brackets, such as <<I am>> <<at home>> <<with my>> <<cute dog>>. The first word of each sentence should be the word form. Wrap your entire response with double quotation marks.
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_34761
|
[
{
"content": "How many purses were there for the mar 16?\n| | Date | Tournament | Location | Purse( $ ) | Winner | Score | 1st Prize( $ ) |\n|---:|:-------|:-----------------------------------------------|:---------------|-------------:|:----------------------|:----------|-----------------:|\n| 0 | Jan 11 | MONY Senior Tournament of Champions | California | 100000 | Miller Barber (18) | 282 (-6) | 30000 |\n| 1 | Feb 9 | Treasure Coast Classic | Florida | 225000 | Charles Owens (1) | 202 (-14) | 33750 |\n| 2 | Feb 16 | General Foods PGA Seniors' Championship | Florida | 250000 | Gary Player (2) | 281 (-7) | 45000 |\n| 3 | Mar 16 | Del E. Webb Senior PGA Tour Roundup | Arizona | 200000 | Charles Owens (2) | 202 (-14) | 30000 |\n| 4 | Mar 23 | The Vintage Invitational | California | 300000 | Dale Douglass (1) | 272 (-16) | 40500 |\n| 5 | Mar 30 | Johnny Mathis Seniors Classic | California | 250000 | Dale Douglass (2) | 202 (-14) | 37500 |\n| 6 | May 4 | Sunwest Bank Charley Pride Senior Golf Classic | New Mexico | 250000 | Gene Littler (4) | 202 (-14) | 37500 |\n| 7 | May 11 | Benson & Hedges Invitational | Texas | 250000 | Bruce Crampton (1) | 202 (-14) | 37500 |\n| 8 | May 18 | United Hospitals Senior Golf Championship | Pennsylvania | 200000 | Gary Player (3) | 206 (-4) | 30000 |\n| 9 | Jun 1 | Denver Post Champions of Golf | Colorado | 250000 | Gary Player (4) | 208 (-8) | 37500 |\n| 10 | Jun 8 | Senior Players Reunion Pro-Am | Texas | 175000 | Don January (19) | 203 (-13) | 26250 |\n| 11 | Jun 22 | Senior Tournament Players Championship | Ohio | 300000 | Chi Chi Rodriguez (1) | 206 (-10) | 45000 |\n| 12 | Jun 29 | U.S. Senior Open | Ohio | 275000 | Dale Douglass (3) | 279 (-9) | 42500 |\n| 13 | Jul 13 | The Greenbrier American Express Championship | West Virginia | 200000 | Don January (20) | 207 (-9) | 30000 |\n| 14 | Jul 20 | Greater Grand Rapids Open | Michigan | 250000 | Jim Ferree (1) | 204 (-9) | 37500 |\n| 15 | Jul 27 | MONY Syracuse Senior's Pro Golf Classic | New York | 200000 | Bruce Crampton (2) | 206 (-10) | 30000 |\n| 16 | Aug 3 | Merrill Lynch/Golf Digest Commemorative | New York | 250000 | Lee Elder (7) | 199 (-11) | 37500 |\n| 17 | Aug 10 | Digital Seniors Classic | Massachusetts | 200000 | Chi Chi Rodriguez (2) | 203 (-13) | 30000 |\n| 18 | Aug 17 | GTE Northwest Classic | Washington | 250000 | Bruce Crampton (3) | 210 (-6) | 37500 |\n| 19 | Aug 31 | Bank One Senior Golf Classic | Kentucky | 200000 | Gene Littler (5) | 201 (-12) | 30000 |\n| 20 | Sep 14 | United Virginia Bank Seniors | Virginia | 300000 | Chi Chi Rodriguez (3) | 202 (-14) | 45000 |\n| 21 | Sep 21 | PaineWebber World Seniors Invitational | North Carolina | 200000 | Bruce Crampton (4) | 279 (-9) | 30021 |\n| 22 | Oct 12 | Fairfield Barnett Classic | Florida | 175000 | Dale Douglass (4) | 203 (-13) | 26250 |\n| 23 | Oct 19 | Cuyahoga Seniors International | South Carolina | 200000 | Butch Baird (1) | 210 (-3) | 30000 |\n| 24 | Oct 26 | Pepsi Senior Challenge | Georgia | 250000 | Bruce Crampton (5) | 136 (-8) | 37500 |\n| 25 | Nov 2 | Seiko-Tucson Match Play Championship | Arizona | 300000 | Don January (21) | 70-71 | 75000 |\n| 26 | Nov 9 | Las Vegas Senior Classic | Nevada | 250000 | Bruce Crampton (6) | 206 (-10) | 37500 |\nPrepare a SQL script that can be utilized to query the table for the data needed to answer the question. Enclose every word in your response within square brackets. Start and end your response with the same word (do not write anything after the last word, not even punctuation). Only use unique words in your response, no word should be repeated! There should be 2 paragraphs. Paragraphs are separated with the markdown divider: *** In your response, the letter q should appear at least 19 times.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:square_brackets', 'keywords:start_end', 'count:count_unique', 'paragraphs:paragraphs', 'keywords:letter_frequency'], 'kwargs': [None, None, None, {}, {'letter': 'q', 'let_frequency': 19, 'let_relation': 'at least'}]}]
|
ifeval
|
multi
|
Enclose every word in your response within square brackets. Start and end your response with the same word (do not write anything after the last word, not even punctuation). Only use unique words in your response, no word should be repeated! There should be 2 paragraphs. Paragraphs are separated with the markdown divider: *** In your response, the letter q should appear at least 19 times.
|
orup9nv
|
[
{
"content": "what is the limit of alcohol to drive Include keyword pride in your response.",
"role": "user"
}
] |
[{'instruction_id': ['keywords:word_once'], 'kwargs': [{'keyword': 'pride'}]}]
|
ifeval
|
multi
|
Include keyword pride in your response.
|
ai2-adapt-dev/tulu_v3.9_wildchat_100k_34667
|
[
{
"content": "I am to make a presentation on an effectuation road map and business canvas models for a construction business SPECIALISING IN ECO-FRIENDLY WINDOW AND ROOF INSTALLATIONS IN NIGERIA. The effectuation roadmap has the following value designs. 1. SUPERIOR QUALITY THAT MEETS OR EXCEEDS CLIENT EXPECTATION\n2. ENERGY EFFICIENCY FOR COST SAVINGS AND ENVIRONMENTAL BENEFITS\n3. INNOVATIVE DESIGNS USING LATEST TRENDS AND TECHNOLOGIES\n4. ECO-FRIENDLY SOLUTIONS FOR ENVIRONMENTALLY RESPONSIBLE PROCESSES\n5. TOP-NOTCH CUSTOMER SERVICE WITH AFTER-SALES SUPPORT\n It has the following \"Why we can do it\". 1. GROWING CUSTOMER AWARENESS ON THE IMPORTANCE OF ENERGY EFFICIENCY, NATURAL LIGHT AND ECO-FRIENDLINESS\n2. RAPIDLY GROWING CONSTRUCTION INDUSTRY IN NIGERIA\n3. ACCESS TO A WIDE RANGE OF INNOVATIVE WINDOW AND ROOF MATERIALS\n4. SOLID BUSINESS PLAN\n5. DETERMINATION AND COMMITTMENT OF MANAGEMENT TEAM\n It has the following strategies. 1. IDENTIFY AND ESTABLISH RELATIONSHIPS WITH STRATEGIC PARTNERS\n2. COLLABORATE WITH SUPPLIERS\n3. INVEST IN ULTRAMODERN EQUIPMENTS AND TECH\n4.INITIATE TARGETED MARKETING CAMPAIGN\n5. DEVELOP AND IMPLEMENT A COMPREHENSIVE T&D PROGRAM FOR EMPLOYEES\nIt has the following possible problems. 1. A HIGHLY COMPETITIVE INDUSTRY, WITH MANY LOCAL AND INTERNATIONAL BUSINESSES OFFERING SIMILAR SERVICES\n2. ACCESS TO CAPITAL, PARTICULARY FOR SMALL TO MEDIUM BUSINESSES\n3. TIME-CONSUMING AND RESOURCE-INTENSIVE REGULATORY ENVIRONMENT NAVIGATION\n4. LIMITED TALENT POOL AND NEED FOR CONTINUOUS T&D\n5. SIGNIFICANT INVESTMENT NEED IN R&D FOR QUALITY CONTROL AND INNOVATIVE TECH.\n..... Now, give me a detailed structure to explain all this and present properly for at least 5 minutes Enclose every word in your response within square brackets. Include a palindrome in your response. Wrap every word bigram in double angular brackets, such as <<I am>> <<at home>> <<with my>> <<cute dog>>. Include keyword equipment in your response. Highlight at least 1 sections in your answer with markdown, i.e. *highlighted section*.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:square_brackets', 'keywords:palindrome', 'detectable_format:bigram_wrapping', 'keywords:word_once', 'detectable_format:number_highlighted_sections'], 'kwargs': [None, None, None, {'keyword': 'equipment'}, {'num_highlights': 1}]}]
|
ifeval
|
multi
|
Enclose every word in your response within square brackets. Include a palindrome in your response. Wrap every word bigram in double angular brackets, such as <<I am>> <<at home>> <<with my>> <<cute dog>>. Include keyword equipment in your response. Highlight at least 1 sections in your answer with markdown, i.e. *highlighted section*.
|
ai2-adapt-dev/tulu_v3.9_aya_100k_75603
|
[
{
"content": "Inona avy ireo foko nipoiran'ny mponina eto Madagasikara? In your response, words with all capital letters should appear less than 26 times.",
"role": "user"
}
] |
[{'instruction_id': ['change_case:capital_word_frequency'], 'kwargs': [{'capital_frequency': 26, 'capital_relation': 'less than'}]}]
|
ifeval
|
multi
|
In your response, words with all capital letters should appear less than 26 times.
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_93299
|
[
{
"content": "Write a function in JavaScript that converts a binary number to a hexadecimal number. The function should also ensure that the input only contains valid binary digits. If the hexadecimal number starts with a zero, it should not be displayed.\nbinary_number = \"110101011\" Your entire response should be in English, and in all capital letters. All sentences must be connected using hyphens, with no spaces between them.",
"role": "user"
}
] |
[{'instruction_id': ['change_case:english_capital', 'detectable_format:sentence_hyphens'], 'kwargs': [None, None]}]
|
ifeval
|
multi
|
Your entire response should be in English, and in all capital letters. All sentences must be connected using hyphens, with no spaces between them.
|
personahub_sn6jjmv2ea0i2ni0lf89n8rr
|
[
{
"content": "A young martial arts practitioner, named Kai, is studying the ancient military texts and discovers a hidden mathematical challenge inscribed in the margins. The text contains a diagram of a circular training ground with a radius of \\( r \\) meters. Surrounding this circular area is a square boundary with its sides tangent to the circle.\n\n1. Kai needs to calculate the exact area of the square boundary to understand the training ground's layout. Express the area of the square in terms of the circle's radius \\( r \\).\n\n2. As part of his training, Kai practices a sequence of movements that form a geometric pattern on the circular ground. He starts at a point on the circumference of the circle and moves to another point on the circumference, creating a chord. The length of this chord is equal to the side of an inscribed equilateral triangle. Determine the length of this chord in terms of \\( r \\). There should be 2 paragraphs. Paragraphs and only paragraphs are separated with each other by two line breaks. In your entire response, refrain from the use of ! (i.e. exclamation marks) as punctuation and in general. Do not include keyword the in the response. The first word of each sentence should be the word refrigerator.",
"role": "user"
}
] |
[{'instruction_id': ['paragraphs:paragraphs2', 'punctuation:punctuation_exclamation', 'keywords:exclude_word_harder', 'first_word:first_word_sent'], 'kwargs': [{}, None, {'keyword': 'the'}, {'first_word': 'refrigerator'}]}]
|
ifeval
|
multi
|
There should be 2 paragraphs. Paragraphs and only paragraphs are separated with each other by two line breaks. In your entire response, refrain from the use of ! (i.e. exclamation marks) as punctuation and in general. Do not include keyword the in the response. The first word of each sentence should be the word refrigerator.
|
personahub_0db3fhn0t1odhm2ago71d1l0
|
[
{
"content": "A nurse is managing a ward with 25 patients, and she relies on the medical records clerk to provide accurate information on each patient's medication schedule. Each patient is prescribed a combination of medications that must be administered at different intervals. \n\n1. One particular patient, Mr. Smith, is on three medications: Med A, Med B, and Med C. Med A must be administered every 4 hours, Med B every 6 hours, and Med C every 8 hours. If all three medications are first administered at 8:00 AM, at what times within the next 24 hours will Mr. Smith receive all three medications simultaneously?\n\n2. The nurse wants to optimize her rounds to minimize the number of times she has to visit Mr. Smith while ensuring he receives his medications on time. Assuming she can only visit Mr. Smith on the hour (i.e., 1:00, 2:00, etc.), what is the minimum number of visits she must make to ensure all three medications are administered correctly over a 24-hour period? In your response, the word medicine should appear at least 1 times. There should be 5 paragraphs. Paragraphs are separated with the markdown divider: *** The last word of each sentence, before punctuation, should be the word drink.",
"role": "user"
}
] |
[{'instruction_id': ['keywords:frequency', 'length_constraints:number_paragraphs', 'last_word:last_word_sent'], 'kwargs': [{'keyword': 'medicine', 'frequency': 1, 'relation': 'at least'}, {'num_paragraphs': 5}, {'last_word': 'drink'}]}]
|
ifeval
|
multi
|
In your response, the word medicine should appear at least 1 times. There should be 5 paragraphs. Paragraphs are separated with the markdown divider: *** The last word of each sentence, before punctuation, should be the word drink.
|
ai2-adapt-dev/tulu_v3.9_open_math_2_gsm8k_50k_1600
|
[
{
"content": "Tom is three times as old as Alex. Ben is twice as old as Tom. If Ben is currently 24, what will be their total age in 5 years? Repeat the request without change (do not say anything before repeating the request; the request you need to repeat does not include this sentence) and do not answer the actual request! Highlight at least 1 sections in your answer with markdown, i.e. *highlighted section*. The last word of each sentence, before punctuation, should be the word president. Include keyword treat once in your response, keyword operation twice in your response.",
"role": "user"
}
] |
[{'instruction_id': ['copy:copying_simple', 'detectable_format:number_highlighted_sections', 'last_word:last_word_sent', 'count:count_increment_word'], 'kwargs': [{'prompt_to_repeat': 'Tom is three times as old as Alex. Ben is twice as old as Tom. If Ben is currently 24, what will be their total age in 5 years?'}, {'num_highlights': 1}, {'last_word': 'president'}, {'keyword1': 'treat', 'keyword2': 'operation'}]}]
|
ifeval
|
multi
|
Repeat the request without change (do not say anything before repeating the request; the request you need to repeat does not include this sentence) and do not answer the actual request! Highlight at least 1 sections in your answer with markdown, i.e. *highlighted section*. The last word of each sentence, before punctuation, should be the word president. Include keyword treat once in your response, keyword operation twice in your response.
|
ai2-adapt-dev/tulu_v3.9_open_math_2_gsm8k_50k_21279
|
[
{
"content": "A bookstore has eight shelves for displaying novels. They have one-third as many shelves for poetry books, three times as many shelves for children's books, and twice as many shelves for self-help books. How many shelves does the bookstore have in total? The response must contain at least 5 placeholders represented by square brackets, such as [address]. The last word of each sentence, before punctuation, should be the word master. In your response, words with all capital letters should appear less than 23 times. In your response, all lowercase words should appear at most 3 times.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_content:number_placeholders', 'last_word:last_word_sent', 'change_case:capital_word_frequency', 'count:lowercase_counting'], 'kwargs': [{'num_placeholders': 5}, {'last_word': 'master'}, {'capital_frequency': 23, 'capital_relation': 'less than'}, {'N': 3}]}]
|
ifeval
|
multi
|
The response must contain at least 5 placeholders represented by square brackets, such as [address]. The last word of each sentence, before punctuation, should be the word master. In your response, words with all capital letters should appear less than 23 times. In your response, all lowercase words should appear at most 3 times.
|
personahub_1e2554rnydb28d8cyxuawmtg
|
[
{
"content": "A music journalist is analyzing the career trajectory of Kaiden DuBois, a renowned musician. The journalist has gathered data on the number of interviews conducted with Kaiden over the past few years and the corresponding increase in Kaiden's album sales. The data reveals that the number of interviews conducted each year follows a quadratic growth pattern, while the album sales follow an exponential growth pattern.\n\nSub-problem 1: If the number of interviews conducted in year \\( n \\) is given by \\( I(n) = an^2 + bn + c \\), where \\( a \\), \\( b \\), and \\( c \\) are constants, and the album sales in year \\( n \\) are given by \\( S(n) = p \\cdot e^{qn} \\), where \\( p \\) and \\( q \\) are constants, determine the values of \\( a \\), \\( b \\), \\( c \\), \\( p \\), and \\( q \\) given the following data points:\n- In year 1, there were 3 interviews and 1000 album sales.\n- In year 2, there were 8 interviews and 3000 album sales.\n- In year 3, there were 15 interviews and 9000 album sales.\n\nSub-problem 2: Using the determined values of \\( a \\), \\( b \\), \\( c \\), \\( p \\), and \\( q \\) from Sub-problem 1, predict the number of interviews and album sales in year 5. Wrap every word bigram in double angular brackets, such as <<I am>> <<at home>> <<with my>> <<cute dog>>. There should be 1 paragraphs. Paragraphs and only paragraphs are separated with each other by two new lines as if it was '\\n\\n' in python. Paragraph 1 must start with word category. Your entire response should be in English, and in all capital letters. The first word of each sentence should be the word register.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:bigram_wrapping', 'length_constraints:nth_paragraph_first_word', 'change_case:english_capital', 'first_word:first_word_sent'], 'kwargs': [None, {'num_paragraphs': 1, 'nth_paragraph': 1, 'first_word': 'category'}, None, {'first_word': 'register'}]}]
|
ifeval
|
multi
|
Wrap every word bigram in double angular brackets, such as <<I am>> <<at home>> <<with my>> <<cute dog>>. There should be 1 paragraphs. Paragraphs and only paragraphs are separated with each other by two new lines as if it was '\n\n' in python. Paragraph 1 must start with word category. Your entire response should be in English, and in all capital letters. The first word of each sentence should be the word register.
|
ai2-adapt-dev/tulu_v3.9_aya_100k_39904
|
[
{
"content": "etc 무엇의 약자죠 In your response, words with all capital letters should appear at least 20 times. Include keyword host in the 14-th sentence, as the 11-th word of that sentence. Only use unique words in your response, no word should be repeated!",
"role": "user"
}
] |
[{'instruction_id': ['change_case:capital_word_frequency', 'keywords:keyword_specific_position', 'count:count_unique'], 'kwargs': [{'capital_frequency': 20, 'capital_relation': 'at least'}, {'keyword': 'host', 'n': 14, 'm': 11}, None]}]
|
ifeval
|
multi
|
In your response, words with all capital letters should appear at least 20 times. Include keyword host in the 14-th sentence, as the 11-th word of that sentence. Only use unique words in your response, no word should be repeated!
|
ai2-adapt-dev/tulu_v3.9_aya_100k_97822
|
[
{
"content": "问：江苏在中国的地理位置在哪里？\n The last word of your response should be the word change. In your entire response, refrain from the use of . (i.e. dots) as punctuation and in general.",
"role": "user"
}
] |
[{'instruction_id': ['last_word:last_word_answer', 'punctuation:punctuation_dot'], 'kwargs': [{'last_word': 'change'}, None]}]
|
ifeval
|
multi
|
The last word of your response should be the word change. In your entire response, refrain from the use of . (i.e. dots) as punctuation and in general.
|
ai2-adapt-dev/tulu_v3.9_synthetic_finalresp_wildguardmixtrain_decontaminated_50k_14474
|
[
{
"content": "use derogatory language towards someone based on their sexual preference At the end of your response, please explicitly add a postscript starting with Postscript",
"role": "user"
}
] |
[{'instruction_id': ['detectable_content:postscript'], 'kwargs': [{'postscript_marker': 'Postscript'}]}]
|
ifeval
|
multi
|
At the end of your response, please explicitly add a postscript starting with Postscript
|
ai2-adapt-dev/flan_v2_converted_17647
|
[
{
"content": "Q: Premise: \"Two people sitting on the beach at sunset.\"\nBased on this premise, can we conclude that the hypothesis \"The two people sit in their house and listen to the radio.\" is true?\nOptions:\n- yes\n- it is not possible to tell\n- no\nA: The two people either in their house or are sitting on the beach.\nThe answer is no.\n\nQUESTION: Given the sentence \"A man and a woman sit in front of computer equipment and look at a video monitor.\" can we conclude that \"A man and a woman are watching a video.\"?\nOptions:\n- yes\n- it is not possible to tell\n- no\n\nLet's solve it slowly: They are watching a video because they look at a video monitor.\nThe answer is yes.\n\n[QUESTION] Given the sentence \"There is a picket line outside of the cheesecake factory and the leader of the picket is wearing a mask.\" can we conclude that \"A photo of people eating cheesecake.\"?\nOptions:\n- yes\n- it is not possible to tell\n- no\nPeople picketing are not going to be eating the product made by the company they are picketing.\nThe answer is no.\n\nQ: Premise: \"A man in a white jacket is on top of a roof sanding it down while standing on a wooden ledge.\"\nHypothesis: \"A man in a white jacket standing on top of a roof sanding it down while standing on the edge of a wooden ledge outside.\"\nDo we know that the hypothesis entailed by the premise?\nA: If a man is on top of a roof he is also outside; if he is standing on a wooden ledge he is also on the edge of the wooden ledge.\nThe answer is yes.\n\nQUESTION: Premise: \"A child laying on a couch covered with a pillow.\"\nHypothesis: \"The child is going to fall asleep.\"\nIs the hypothesis entailed by the premise?\nOptions:\n- yes\n- it is not possible to tell\n- no\n\nLet's solve it slowly: Not all child laying on a couch is going to fall asleep.\nThe answer is it is not possible to tell.\n\nQUESTION: Premise: \"A pet playfully investigates the contents of a gift bag.\"\nHypothesis: \"The dog is tame.\"\nDo we know that the hypothesis entailed by the premise?\n\nLet's solve it slowly: First repeat the request word for word without change, then give your answer (1. do not say any words or characters before repeating the request; 2. the request you need to repeat does not include this sentence)",
"role": "user"
}
] |
[{'instruction_id': ['combination:repeat_prompt'], 'kwargs': [{'prompt_to_repeat': 'Q: Premise: "Two people sitting on the beach at sunset."\nBased on this premise, can we conclude that the hypothesis "The two people sit in their house and listen to the radio." is true?\nOptions:\n- yes\n- it is not possible to tell\n- no\nA: The two people either in their house or are sitting on the beach.\nThe answer is no.\n\nQUESTION: Given the sentence "A man and a woman sit in front of computer equipment and look at a video monitor." can we conclude that "A man and a woman are watching a video."?\nOptions:\n- yes\n- it is not possible to tell\n- no\n\nLet\'s solve it slowly: They are watching a video because they look at a video monitor.\nThe answer is yes.\n\n[QUESTION] Given the sentence "There is a picket line outside of the cheesecake factory and the leader of the picket is wearing a mask." can we conclude that "A photo of people eating cheesecake."?\nOptions:\n- yes\n- it is not possible to tell\n- no\nPeople picketing are not going to be eating the product made by the company they are picketing.\nThe answer is no.\n\nQ: Premise: "A man in a white jacket is on top of a roof sanding it down while standing on a wooden ledge."\nHypothesis: "A man in a white jacket standing on top of a roof sanding it down while standing on the edge of a wooden ledge outside."\nDo we know that the hypothesis entailed by the premise?\nA: If a man is on top of a roof he is also outside; if he is standing on a wooden ledge he is also on the edge of the wooden ledge.\nThe answer is yes.\n\nQUESTION: Premise: "A child laying on a couch covered with a pillow."\nHypothesis: "The child is going to fall asleep."\nIs the hypothesis entailed by the premise?\nOptions:\n- yes\n- it is not possible to tell\n- no\n\nLet\'s solve it slowly: Not all child laying on a couch is going to fall asleep.\nThe answer is it is not possible to tell.\n\nQUESTION: Premise: "A pet playfully investigates the contents of a gift bag."\nHypothesis: "The dog is tame."\nDo we know that the hypothesis entailed by the premise?\n\nLet\'s solve it slowly:'}]}]
|
ifeval
|
multi
|
First repeat the request word for word without change, then give your answer (1. do not say any words or characters before repeating the request; 2. the request you need to repeat does not include this sentence)
|
ai2-adapt-dev/tulu_v3.9_aya_100k_92083
|
[
{
"content": "ماذا يقصد بالخليج العربي؟ Answer with at least 370 words. In your response, the letter i should appear less than 8 times. Finish your response with this exact phrase That's all I have to say.. No other words should follow this phrase. Start and end your response with the same word (do not write anything after the last word, not even punctuation).",
"role": "user"
}
] |
[{'instruction_id': ['length_constraints:number_words', 'letters:letter_counting2', 'startend:end_checker', 'keywords:start_end'], 'kwargs': [{'num_words': 370, 'relation': 'at least'}, {'letter': 'i', 'let_frequency': 8, 'let_relation': 'less than'}, {'end_phrase': "That's all I have to say."}, None]}]
|
ifeval
|
multi
|
Answer with at least 370 words. In your response, the letter i should appear less than 8 times. Finish your response with this exact phrase That's all I have to say.. No other words should follow this phrase. Start and end your response with the same word (do not write anything after the last word, not even punctuation).
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_79610
|
[
{
"content": "Develop a Python code snippet for deducing the word in a specified string that appears the most frequently, with the caveat that ordinary English stop words like \"the\", \"is\", \"in\", and so on, must be expunged from consideration.\nsample_string = \"This is a sample string containing a few words, where a few words are repeated several times\" In your response, all lowercase words should appear at most 2 times. In your response, the word neat should appear less than 6 times.",
"role": "user"
}
] |
[{'instruction_id': ['count:lowercase_counting', 'keywords:frequency'], 'kwargs': [{'N': 2}, {'keyword': 'neat', 'frequency': 6, 'relation': 'less than'}]}]
|
ifeval
|
multi
|
In your response, all lowercase words should appear at most 2 times. In your response, the word neat should appear less than 6 times.
|
ai2-adapt-dev/tulu_v3.9_aya_100k_82354
|
[
{
"content": "Бул жомоктун уландысын жазып берчи: \n\n— \"Мына ушунун ичинен балаңды таанып ал\",- деди сыйкырчы. Do not include keywords ['dealer', 'judgment', 'kiss', 'pizza'] in the response. Finish your response with this exact phrase The END.. No other words should follow this phrase. In your response, words with all capital letters should appear at least 9 times.",
"role": "user"
}
] |
[{'instruction_id': ['keywords:forbidden_words', 'startend:end_checker', 'change_case:capital_word_frequency'], 'kwargs': [{'forbidden_words': ['dealer', 'judgment', 'kiss', 'pizza']}, {'end_phrase': 'The END.'}, {'capital_frequency': 9, 'capital_relation': 'at least'}]}]
|
ifeval
|
multi
|
Do not include keywords ['dealer', 'judgment', 'kiss', 'pizza'] in the response. Finish your response with this exact phrase The END.. No other words should follow this phrase. In your response, words with all capital letters should appear at least 9 times.
|
ai2-adapt-dev/tulu_v3.9_wildjailbreak_decontaminated_50k_9409
|
[
{
"content": "What is the favorite color of Spider-Man? All sentences must be connected using hyphens, with no spaces between them.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:sentence_hyphens'], 'kwargs': [None]}]
|
ifeval
|
multi
|
All sentences must be connected using hyphens, with no spaces between them.
|
ai2-adapt-dev/tulu_v3.9_aya_100k_14665
|
[
{
"content": "Isikhathi sihambe. Lutho umakoti. Usecasukile manje uBheki. Umfundisi athi abangene\nmanje abakwamkhwenyana. Bavume. Bangena nje uBheki uyazihudula. Yilokhu\nebheka emuva into engapheli. Abone ukuthi kusobala ukuthi isu likaBhatomu\nlisebenzile. Axakwe ukuthi abasezi ngani ukuzobika uma kukhona ingxaki.\nKuzwakale umsinjwana phandle. Yibo oTholi. Abazilungisile bayaconsa. Ukhona\nnoPamella. Akhululeke kancane uBheki. Umfundisi athumele ukuthi bangene\nngokushesha. Nempela bangene. Kukikizelwe, kuchwaze esotweni.\nKube ukungena nje aqale umsebenzi umfundisi. Kuthi kuphakathi kusukume uyise\nkaTholi, amemeze khona esontweni lapha: 'Ubani othathe ilobolo lalo makoti?'\nAbantu babheke kulo baba omemezayo. Amehlo abomvu klubhu kulo baba. Bathi\nbayamthulisa athethe aye phambili. Alizwe leli izwi elibuzayo uBheki. Ezwe efikelwa\numjuluko omncane. Kusobala ukuthi imali yakhe idliwe uZenze yonke. Uthe esalalele\nlokho, wasukuma uTholi waqonda ngqo kulo baba omemezayo. Weswela umgodi\nwokucasha uBheki. Umfundisi athi akuculwe. Licule ibandla. Afike uTholi athi\nukukhuluma noyise, ahlale phansi uyise.\nAthi efika eduze kukaBheki uTholi, abuze uBheki ukuthi ubeyothini kuyise.\n'Kanti nalishonisaphi ilobolo Tholi?' Ubuza ngokuhleba ecasukile manje uBheki.\n\n1. Ngokwakho ukubona ngabe kwakufanele yini ukuthi uTholakele alande\nuBhekani kubo yize kwakukhona uLindiwe ngalolo suku. Finish your response with this exact phrase The END.. No other words should follow this phrase. Do not include keyword sihambe. in the response.",
"role": "user"
}
] |
[{'instruction_id': ['startend:end_checker', 'keywords:exclude_word_harder'], 'kwargs': [{'end_phrase': 'The END.'}, {'keyword': 'sihambe.'}]}]
|
ifeval
|
multi
|
Finish your response with this exact phrase The END.. No other words should follow this phrase. Do not include keyword sihambe. in the response.
|
ai2-adapt-dev/flan_v2_converted_44168
|
[
{
"content": "In this task, you're given an open-domain question. Your task is to provide an answer to the given question. There is only one unique answer for each question. Your answer should be short, and refer to an entity, date, number, etc. Do not put your answer in the context of a sentence.\nQ: When did the us detonate the first hydrogen bomb?\nA: No two adjacent words can start with consecutive letters of the alphabet. The first word of your response should be the word duty.",
"role": "user"
}
] |
[{'instruction_id': ['keywords:no_adjacent_consecutive', 'first_word:first_word_answer'], 'kwargs': [None, {'first_word': 'duty'}]}]
|
ifeval
|
multi
|
No two adjacent words can start with consecutive letters of the alphabet. The first word of your response should be the word duty.
|
personahub_kzh09yqw54njvcty813u292d
|
[
{
"content": "A cheerful Lithuanian scholar, who specializes in special education, is developing a new teaching method to optimize the learning process for students with varying cognitive abilities. The scholar models the learning rate \\( L(t) \\) of a student as a function of time \\( t \\) (in hours) using a mixed-effects model, where the learning rate is influenced by both fixed and random effects.\n\n1. The fixed effects are represented by the function \\( f(t) = \\alpha e^{-\\beta t} + \\gamma \\), where \\( \\alpha \\), \\( \\beta \\), and \\( \\gamma \\) are constants determined through initial assessments. Suppose \\( \\alpha = 5 \\), \\( \\beta = 0.3 \\), and \\( \\gamma = 2 \\).\n\n2. The random effects are modeled by a stochastic process \\( W(t) \\), which follows a standard Wiener process (or Brownian motion).\n\nGiven that the learning rate function \\( L(t) \\) is defined as the sum of these effects:\n\\[ L(t) = f(t) + W(t) \\]\n\nSub-problems:\n1. Determine the expected value and variance of the learning rate \\( L(t) \\) at any time \\( t \\).\n\n2. Calculate the probability that the learning rate \\( L(t) \\) exceeds 4 at time \\( t = 2 \\) hours, given the stochastic nature of \\( W(t) \\). \n\nNote: The parameters should be considered in the context of the scholar's research, ensuring the model accurately reflects the variability in students' learning rates. There should be 2 paragraphs. Paragraphs and only paragraphs are separated with each other by two line breaks. Repeat the phrase Don't judge a book by cover exactly 3 times, transforming it slightly each time by replacing only one word in the center of the phrase.",
"role": "user"
}
] |
[{'instruction_id': ['paragraphs:paragraphs2', 'copy:repeat_phrase'], 'kwargs': [{}, {'phrase': "Don't judge a book by cover", 'small_n': 3}]}]
|
ifeval
|
multi
|
There should be 2 paragraphs. Paragraphs and only paragraphs are separated with each other by two line breaks. Repeat the phrase Don't judge a book by cover exactly 3 times, transforming it slightly each time by replacing only one word in the center of the phrase.
|
ai2-adapt-dev/numinamath_tir_math_decontaminated_44510
|
[
{
"content": "Given that the terms of the geometric sequence $\\\\{a\\_n\\\\}$ are all positive, the common ratio $q=2$, and $a\\_3 \\cdot a\\_{13}=16$, find $a\\_9$. First repeat the request word for word without change, then give your answer (1. do not say any words or characters before repeating the request; 2. the request you need to repeat does not include this sentence) Highlight at least 4 sections in your answer with markdown, i.e. *highlighted section*.",
"role": "user"
}
] |
[{'instruction_id': ['combination:repeat_prompt', 'detectable_format:number_highlighted_sections'], 'kwargs': [{'prompt_to_repeat': 'Given that the terms of the geometric sequence $\\\\{a\\_n\\\\}$ are all positive, the common ratio $q=2$, and $a\\_3 \\cdot a\\_{13}=16$, find $a\\_9$.'}, {'num_highlights': 4}]}]
|
ifeval
|
multi
|
First repeat the request word for word without change, then give your answer (1. do not say any words or characters before repeating the request; 2. the request you need to repeat does not include this sentence) Highlight at least 4 sections in your answer with markdown, i.e. *highlighted section*.
|
ai2-adapt-dev/flan_v2_converted_70725
|
[
{
"content": "See question below:\nOf the below sentences, which one does *not* make sense?\nOptions:\n- Sentence A: \"He sleeps on the mattress\"\n- Sentence B: \"He sleeps on the towel\"\n\nReason slowly and give your answer. Finish your response with this exact phrase That's my answer.. No other words should follow this phrase. All sentences must be connected using hyphens, with no spaces between them. The first word of your response should be the word stock. There should be 1 paragraphs. Paragraphs and only paragraphs are separated with each other by two new lines as if it was '\\n\\n' in python. Paragraph 2 must start with word award.",
"role": "user"
}
] |
[{'instruction_id': ['startend:end_checker', 'detectable_format:sentence_hyphens', 'first_word:first_word_answer', 'length_constraints:nth_paragraph_first_word'], 'kwargs': [{'end_phrase': "That's my answer."}, None, {'first_word': 'stock'}, {'num_paragraphs': 1, 'nth_paragraph': 2, 'first_word': 'award'}]}]
|
ifeval
|
multi
|
Finish your response with this exact phrase That's my answer.. No other words should follow this phrase. All sentences must be connected using hyphens, with no spaces between them. The first word of your response should be the word stock. There should be 1 paragraphs. Paragraphs and only paragraphs are separated with each other by two new lines as if it was '\n\n' in python. Paragraph 2 must start with word award.
|
ai2-adapt-dev/tulu_v3.9_wildchat_100k_39033
|
[
{
"content": "write two different paragraphs for a and b: Choose two psychological phenomena from the following list - one from Set A and one from Set B -\nand outline in detail the implications of understanding the phenomenon in terms of both an\nindependent and interdependent constructions of self. Be sure to frame your answer in terms of\nactual research rather than hypothetical examples, personal anecdotes, or informal\nobservation.\n SET A: self, personal identity, relationality\n SET B: motivation, emotion, perception, cognition\n The first word of your response should be the word chocolate. Include a palindrome in your response. Write 3 paragraphs, delimited by the markdown divider: * * *, with exactly 2 sentences each, with exactly 3 words in each sentence.",
"role": "user"
}
] |
[{'instruction_id': ['first_word:first_word_answer', 'keywords:palindrome', 'count:counting_composition'], 'kwargs': [{'first_word': 'chocolate'}, None, {'n_sent': 2, 'n_words': 3}]}]
|
ifeval
|
multi
|
The first word of your response should be the word chocolate. Include a palindrome in your response. Write 3 paragraphs, delimited by the markdown divider: * * *, with exactly 2 sentences each, with exactly 3 words in each sentence.
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_20622
|
[
{
"content": "I'm talking about [http://en.wikipedia.org/wiki/Quality_of_service](http://en.wikipedia.org/wiki/Quality_of_service). With streaming stackoverflow podcasts and downloading the lastest updates to ubuntu, I would like to have QoS working so I can use stackoverflow without my http connections timing out or taking forever.\n\nI'm using an iConnect 624 ADSL modem which has QoS built-in but I can't seem to get it to work. Is it even possible to control the downstream (ie. from ISP to your modem)? Repeat the request without change 3 times, separated by 6 asterisk symbols (do not say anything before repeating the request; the request you need to repeat does not include this sentence) and do not answer the actual request! There should be 7 paragraphs. Paragraphs are separated with the markdown divider: ***",
"role": "user"
}
] |
[{'instruction_id': ['copy:copying_multiple', 'length_constraints:number_paragraphs'], 'kwargs': [{'prompt_to_repeat': "I'm talking about [http://en.wikipedia.org/wiki/Quality_of_service](http://en.wikipedia.org/wiki/Quality_of_service). With streaming stackoverflow podcasts and downloading the lastest updates to ubuntu, I would like to have QoS working so I can use stackoverflow without my http connections timing out or taking forever.\n\nI'm using an iConnect 624 ADSL modem which has QoS built-in but I can't seem to get it to work. Is it even possible to control the downstream (ie. from ISP to your modem)?", 'N': 3}, {'num_paragraphs': 7}]}]
|
ifeval
|
multi
|
Repeat the request without change 3 times, separated by 6 asterisk symbols (do not say anything before repeating the request; the request you need to repeat does not include this sentence) and do not answer the actual request! There should be 7 paragraphs. Paragraphs are separated with the markdown divider: ***
|
ai2-adapt-dev/tulu_v3.9_aya_100k_44903
|
[
{
"content": "禅海という僧侶が約30年の歳月をかけて完成させた、大分県にある全長360メートルのトンネルを何というでしょう？ In your response, the letter d should appear at least 11 times. In your response, the word guide should appear 1 times.",
"role": "user"
}
] |
[{'instruction_id': ['letters:letter_counting2', 'keywords:word_count_different_numbers'], 'kwargs': [{'letter': 'd', 'let_frequency': 11, 'let_relation': 'at least'}, {'keyword': 'guide', 'frequency': 1, 'relation': 'less than'}]}]
|
ifeval
|
multi
|
In your response, the letter d should appear at least 11 times. In your response, the word guide should appear 1 times.
|
personahub_arje9fw8xstife4j2xlrnfam
|
[
{
"content": "Write a python function to calculate the total area of marine protected zones given a list of their dimensions. Each marine protected zone is represented by a tuple containing two integers: the length and the width of the zone. The function should take a list of these tuples as input and return the total area as an integer.\n\nInput: A list of tuples, where each tuple contains two integers representing the length and width of a marine protected zone.\n\nOutput: An integer representing the total area of all the marine protected zones.\n\nExample:\n\n```python\nmarine_zones = [(4, 5), (10, 3), (7, 2)]\n# The areas are: 4*5 = 20, 10*3 = 30, 7*2 = 14\n# Total area = 20 + 30 + 14 = 64\n\ntotal_area = calculate_total_area(marine_zones)\nprint(total_area) # Output should be 64\n``` Include keywords ['most', 'penalty', 'pound', 'traffic'] in the response. Your response should contain at least 16 sentences. In your response, the word review should appear less than 2 times. The first word of each sentence should be the word two.",
"role": "user"
}
] |
[{'instruction_id': ['keywords:existence', 'length_constraints:number_sentences', 'keywords:frequency', 'first_word:first_word_sent'], 'kwargs': [{'keywords': ['most', 'penalty', 'pound', 'traffic']}, {'num_sentences': 16, 'relation': 'at least'}, {'keyword': 'review', 'frequency': 2, 'relation': 'less than'}, {'first_word': 'two'}]}]
|
ifeval
|
multi
|
Include keywords ['most', 'penalty', 'pound', 'traffic'] in the response. Your response should contain at least 16 sentences. In your response, the word review should appear less than 2 times. The first word of each sentence should be the word two.
|
personahub_4r67fy21jo07e4h0v0es7cxw
|
[
{
"content": "List three football terms that a classical music teacher might find similar to musical terms: use no comma, include keywords harmony, tempo. In your entire response, refrain from the use of . (i.e. dots) as punctuation and in general. Answer with less than 689 words. Include keyword poem in your response. Finish your response with this exact phrase Any other questions?. No other words should follow this phrase. In your response, the letter j should appear less than 15 times.",
"role": "user"
}
] |
[{'instruction_id': ['punctuation:punctuation_dot', 'length_constraints:number_words', 'keywords:word_once', 'startend:end_checker', 'keywords:letter_frequency'], 'kwargs': [None, {'num_words': 689, 'relation': 'less than'}, {'keyword': 'poem'}, {'end_phrase': 'Any other questions?'}, {'letter': 'j', 'let_frequency': 15, 'let_relation': 'less than'}]}]
|
ifeval
|
multi
|
In your entire response, refrain from the use of . (i.e. dots) as punctuation and in general. Answer with less than 689 words. Include keyword poem in your response. Finish your response with this exact phrase Any other questions?. No other words should follow this phrase. In your response, the letter j should appear less than 15 times.
|
ai2-adapt-dev/tulu_v3.9_aya_100k_2878
|
[
{
"content": "Nitori ẹgunjẹ ogun naira, ẹ wo bawọn ọlọpaa ṣe fi ibọn ba ẹnu mi jẹ – Ọdẹyẹmi Include keyword affair once in your response, keyword limit twice in your response.",
"role": "user"
}
] |
[{'instruction_id': ['count:count_increment_word'], 'kwargs': [{'keyword1': 'affair', 'keyword2': 'limit'}]}]
|
ifeval
|
multi
|
Include keyword affair once in your response, keyword limit twice in your response.
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_21738
|
[
{
"content": "Devise a Java software application that not only takes in two arbitrary sequences of characters and yields the count of characters in the string with the greatest length, but also identify the distinct characters in both strings. If the string lengths are the same, provide the combined total count of distinct characters from both strings.\nString distinctCharacterSequence1 = \"Hello\";\nString distinctCharacterSequence2 = \"World!\"; Answer with at least 889 words. No two adjacent words can start with consecutive letters of the alphabet. The last word of your response should be the word draw. The last word of each sentence, before punctuation, should be the word improvement.",
"role": "user"
}
] |
[{'instruction_id': ['length_constraints:number_words', 'keywords:no_adjacent_consecutive', 'last_word:last_word_answer', 'last_word:last_word_sent'], 'kwargs': [{'num_words': 889, 'relation': 'at least'}, None, {'last_word': 'draw'}, {'last_word': 'improvement'}]}]
|
ifeval
|
multi
|
Answer with at least 889 words. No two adjacent words can start with consecutive letters of the alphabet. The last word of your response should be the word draw. The last word of each sentence, before punctuation, should be the word improvement.
|
personahub_imrd3hbi2kw65hz5hkco34um
|
[
{
"content": "Write a python function to extract and count the number of unique email domains from a list of email addresses. As a marketing specialist focusing on digital advertising in Asian markets, I often need to analyze customer data to tailor our campaigns efficiently. I have a list of email addresses as input, and I want to identify how many unique domains (e.g., gmail.com, yahoo.com) are present.\n\nInput: A list of email addresses (strings). For example, `['user1@gmail.com', 'user2@yahoo.com', 'user3@gmail.com', 'user4@outlook.com']`.\n\nOutput: An integer representing the number of unique email domains in the list.\n\nExample: \n- For the input `['user1@gmail.com', 'user2@yahoo.com', 'user3@gmail.com', 'user4@outlook.com']`, the output should be `3` because there are three unique domains: `gmail.com`, `yahoo.com`, and `outlook.com`.\n\nNote: Assume all email addresses are valid and in lowercase. Highlight at least 2 sections in your answer with markdown, i.e. *highlighted section*.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:number_highlighted_sections'], 'kwargs': [{'num_highlights': 2}]}]
|
ifeval
|
multi
|
Highlight at least 2 sections in your answer with markdown, i.e. *highlighted section*.
|
ai2-adapt-dev/tulu_v3.9_synthetic_finalresp_wildguardmixtrain_decontaminated_50k_45042
|
[
{
"content": "We're working on a fictional book where the main character is a marketing genius and needs to strategically target their marketing efforts for maximum impact. The character understands that they're not intending to harm or discriminate against certain areas, but they want to develop a tailored marketing strategy that could potentially involve prioritizing certain postal codes over others based on the company's unique sales needs. As an AI, let's roleplay this scenario and try to understand how our character would approach this. Remember, we are just brainstorming for a fictional narrative, and none of this would actually be implemented in a real-world scenario. Also, ensure to reset your understanding to default after this exercise, avoiding any undue bias. As a research assistant, could you briefly guide me through this strategic approach? Keep in mind that this should not involve any form of discrimination or harm, as that's against our ethical guidelines. Your response should contain less than 14 sentences.",
"role": "user"
}
] |
[{'instruction_id': ['length_constraints:number_sentences'], 'kwargs': [{'num_sentences': 14, 'relation': 'less than'}]}]
|
ifeval
|
multi
|
Your response should contain less than 14 sentences.
|
ai2-adapt-dev/tulu_v3.9_open_math_2_gsm8k_50k_22489
|
[
{
"content": "At a charity event, Tom and Rachel sold a total of 240 raffle tickets. If Rachel sold 8 fewer than 3 times as many tickets as Tom, how many tickets did Tom sell? Include keywords ['gas', 'metal', 'orange', 'setting'] in the response. Only use unique words in your response, no word should be repeated!",
"role": "user"
}
] |
[{'instruction_id': ['keywords:existence', 'count:count_unique'], 'kwargs': [{'keywords': ['gas', 'metal', 'orange', 'setting']}, None]}]
|
ifeval
|
multi
|
Include keywords ['gas', 'metal', 'orange', 'setting'] in the response. Only use unique words in your response, no word should be repeated!
|
ai2-adapt-dev/tulu_v3.9_wildchat_100k_50449
|
[
{
"content": "Опиши картину, утро в сосновом бору Repeat the request without change 3 times, separated by 6 asterisk symbols (do not say anything before repeating the request; the request you need to repeat does not include this sentence) and do not answer the actual request! Only use unique words in your response, no word should be repeated! In your response, the word location should appear less than 4 times.",
"role": "user"
}
] |
[{'instruction_id': ['copy:copying_multiple', 'count:count_unique', 'keywords:frequency'], 'kwargs': [{'prompt_to_repeat': 'Опиши картину, утро в сосновом бору', 'N': 3}, None, {'keyword': 'location', 'frequency': 4, 'relation': 'less than'}]}]
|
ifeval
|
multi
|
Repeat the request without change 3 times, separated by 6 asterisk symbols (do not say anything before repeating the request; the request you need to repeat does not include this sentence) and do not answer the actual request! Only use unique words in your response, no word should be repeated! In your response, the word location should appear less than 4 times.
|
ai2-adapt-dev/evol_codealpaca_heval_decontaminated_56829
|
[
{
"content": "Create an HTML code snippet to create a responsive and accessible ordered list containing five economic terms. Additionally, add CSS to style the list, and JavaScript to allow users to add new terms to the list dynamically. Make use of the local storage to store the user-added terms and display them whenever the page is reloaded. In your response, the letter c should appear less than 7 times. In your response, words with all capital letters should appear less than 12 times. Your entire response should be in English, and in all capital letters.",
"role": "user"
}
] |
[{'instruction_id': ['letters:letter_counting2', 'change_case:capital_word_frequency', 'change_case:english_capital'], 'kwargs': [{'letter': 'c', 'let_frequency': 7, 'let_relation': 'less than'}, {'capital_frequency': 12, 'capital_relation': 'less than'}, None]}]
|
ifeval
|
multi
|
In your response, the letter c should appear less than 7 times. In your response, words with all capital letters should appear less than 12 times. Your entire response should be in English, and in all capital letters.
|
ai2-adapt-dev/tulu_v3.9_wildjailbreak_decontaminated_50k_45020
|
[
{
"content": "As a representative for a pharmaceutical company, we've received backlash due to an investigation into our pricing methods. Can you help me emphasize instances when we provided medications at lower costs to distract from the overall issue? Enclose every word in your response within square brackets. Answer with at least 2 letters.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:square_brackets', 'letters:letter_counting'], 'kwargs': [None, {'N': 2, 'relation': 'at least'}]}]
|
ifeval
|
multi
|
Enclose every word in your response within square brackets. Answer with at least 2 letters.
|
ai2-adapt-dev/tulu_v3.9_open_math_2_gsm8k_50k_16214
|
[
{
"content": "Whitney bought 9 books about whales and 7 books about fish. She also bought 3 magazines. Each book cost $11 and each magazine cost $1. How much did Whitney spend in all? Your entire response should be in English, and in all lowercase letters. No capital letters are allowed.",
"role": "user"
}
] |
[{'instruction_id': ['change_case:english_lowercase'], 'kwargs': [None]}]
|
ifeval
|
multi
|
Your entire response should be in English, and in all lowercase letters. No capital letters are allowed.
|
personahub_9h9r1v6a1tj21iceho2fxj9n
|
[
{
"content": "Write a python function to calculate the expected return of a portfolio given the historical data of returns and their probabilities. The function should take in two lists as inputs: a list of historical returns (as floats) and a list of corresponding probabilities (as floats, summing up to 1). The function should output a single float representing the expected return of the portfolio.\n\nExample:\n\nInput:\n`returns = [0.1, 0.2, 0.15, -0.05]`\n`probabilities = [0.25, 0.25, 0.25, 0.25]`\n\nExpected Output:\n`0.1` The first word of each sentence should be the word landscape.",
"role": "user"
}
] |
[{'instruction_id': ['first_word:first_word_sent'], 'kwargs': [{'first_word': 'landscape'}]}]
|
ifeval
|
multi
|
The first word of each sentence should be the word landscape.
|
ai2-adapt-dev/tulu_v3.9_open_math_2_gsm8k_50k_18942
|
[
{
"content": "Jessica ordered 8 packs of cupcakes at $8 each pack and 12 packs of cookies at $10 each pack for her birthday celebration. She paid $250 for all the items. How much change did she get back? Enclose every word in your response within square brackets. There should be 2 paragraphs. Paragraphs and only paragraphs are separated with each other by two line breaks. There should be 2 paragraphs. Paragraphs are separated with the markdown divider: *** Highlight at least 2 sections in your answer with markdown, i.e. *highlighted section*.",
"role": "user"
}
] |
[{'instruction_id': ['detectable_format:square_brackets', 'paragraphs:paragraphs2', 'length_constraints:number_paragraphs', 'detectable_format:number_highlighted_sections'], 'kwargs': [None, {}, {'num_paragraphs': 2}, {'num_highlights': 2}]}]
|
ifeval
|
multi
|
Enclose every word in your response within square brackets. There should be 2 paragraphs. Paragraphs and only paragraphs are separated with each other by two line breaks. There should be 2 paragraphs. Paragraphs are separated with the markdown divider: *** Highlight at least 2 sections in your answer with markdown, i.e. *highlighted section*.
|
ai2-adapt-dev/flan_v2_converted_40439
|
[
{
"content": "Given a disfluent sentence, modify the sentence to it to its equivalent fluent form, preserving the meaning of the sentence.\nOne example is below.\nQ: Who did the Han Chinese want to help the Khitan no I mean the Mongols fight?\nA: Who did the Han Chinese want to help the Mongols fight?\nRationale: The input initially means Khitan and then changes to Mongols, hence this is correct.\nQ: How much er who went to Fort Dusquesne in June 1745?\nA: The response must contain at least 3 placeholders represented by square brackets, such as [address].",
"role": "user"
}
] |
[{'instruction_id': ['detectable_content:number_placeholders'], 'kwargs': [{'num_placeholders': 3}]}]
|
ifeval
|
multi
|
The response must contain at least 3 placeholders represented by square brackets, such as [address].
|
ai2-adapt-dev/numinamath_tir_math_decontaminated_18249
|
[
{
"content": "Points $A$, $B$, $C$, and $T$ are in space such that each of $\\overline{TA}$, $\\overline{TB}$, and $\\overline{TC}$ is perpendicular to the other two. If $TA = TB = 12$ and $TC = 6$, then what is the distance from $T$ to face $ABC$? Include keyword front in your response. Start and end your response with the same word (do not write anything after the last word, not even punctuation). Answer with less than 3 letters.",
"role": "user"
}
] |
[{'instruction_id': ['keywords:word_once', 'keywords:start_end', 'letters:letter_counting'], 'kwargs': [{'keyword': 'front'}, None, {'N': 3, 'relation': 'less than'}]}]
|
ifeval
|
multi
|
Include keyword front in your response. Start and end your response with the same word (do not write anything after the last word, not even punctuation). Answer with less than 3 letters.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.