id stringlengths 13 36 | type stringclasses 12
values | constraint stringlengths 17 179 | verifier_type stringclasses 2
values | description stringclasses 13
values | content_template bool 2
classes | verifier_prompt stringclasses 7
values | quality int64 2 5 | frequency int64 1 5 |
|---|---|---|---|---|---|---|---|---|
hard-length_count-000 | length_count | Ensure the response is exact 300 words. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-001 | length_count | Ensure the response is between 400 to 500 words. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-002 | length_count | Response should be about 150 words. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-003 | length_count | Response should be only one paragraph. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-004 | length_count | Limit the first paragraph less than 150 words. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-005 | length_count | Response must contain exact 3 paragraphs. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-006 | length_count | Response must be more than 3 paragraphs. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-007 | length_count | Respond must be less than 5 sentences. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-008 | length_count | Each sentence must contain 10–12 words, inclusive. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-009 | length_count | No word in the response may exceed 6 letters. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-010 | length_count | Ensure the response has exactly 2 sentences per paragraph, with 5 paragraphs total. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-011 | length_count | The response satisfies: the first sentence has 5 words, the second 6, the third 7, and so on until 10. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-012 | length_count | Every sentence in the response must end with exactly 8 words. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-length_count-013 | length_count | The last sentence in the response must contain exactly 4 words. | hard | Strictly or loosely limiting the length of the paragraphs, sentences, words, etc. | false | - If the constraint requires exact words/sentences/paragraphs/length/count, you should strictly verify the quantity.
- If the constraint gives the exact interval of the quantity, you should strictly verify the quantity within the interval.
- If the constraint gives an approximate quantity, you do an interval check. For... | 5 | 5 |
hard-layout-000 | layout | The entire response should be in capital letters. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-001 | layout | First word in a paragraph should be capital. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-002 | layout | All titles and subtitles should be capital. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-003 | layout | Use two empty lines between paragraphs. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-004 | layout | Each line should be 50 characters; use hyphen to break a word if it exceeds 50 characters. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-005 | layout | Use underscore to break letters. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-006 | layout | Make the first sentence in each paragraph be capital. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-007 | layout | Make the first sentence in each paragraph be bold using Markdown **bold**. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-008 | layout | Right-align every line by padding spaces to 60 characters. Each line maximal 80 characters | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-009 | layout | Indent the first line of each paragraph with 4 spaces; others unindented. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-010 | layout | Make each line maximal 60 characters and prefix each line with a three-digit line number and a space (e.g., 001 ) | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-011 | layout | Each line should be 78 characters; the last line should be greater than half of the linewidth. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-layout-012 | layout | Ensure every equation starts on a new line. | hard | Adding constraints on the layout, e.g., letters, space, indent, lines, margin, etc. | false | 5 | 4 | |
hard-file_format-000 | file_format | Ensure the response is JSON format. Wrapped in ```json``` | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-001 | file_format | Wrap the response into an email with subject, to, content (salutation, main body, closing, signature). | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-002 | file_format | Ensure the response is in Latex. Wrapped in ```latex``` | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-003 | file_format | The response should use # and ## markdown format. | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-004 | file_format | Wrap the response with 4 markdown bullets. | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-005 | file_format | Ensure the response is in a Latex table. | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-006 | file_format | Emphasize 4 key words with markdown format. | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-007 | file_format | Wrap the response with 4 points using Latex itemize: \begin{itemize}...\end{itemize} | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-008 | file_format | Include 2 Level 1 markdown heading and 3 Level 2 markdown heading under each Level 1 heading. | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-009 | file_format | Use emoji numbers (e.g.1️⃣2️⃣) as bullets. | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-file_format-010 | file_format | Make the response follow the template: \n# [TITLE]\n[DATE in yyyy/mm/dd] by [AUTHOR]\n[MAIN BODY]\n## Reference\n[REFERENCE] | hard | Responding with some self-defined or wellknown format or template. | false | 5 | 4 | |
hard-linguistic_format-000 | linguistic_format | Do not use adverb. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-001 | linguistic_format | All sentences are in passive voice. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-002 | linguistic_format | The response should use past tense. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-003 | linguistic_format | Contain at least three relative clauses. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-004 | linguistic_format | Use no more than 4 transition words. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-005 | linguistic_format | Only use English. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-006 | linguistic_format | The response should only use the first half of alphabet (A-M). | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-007 | linguistic_format | The response should only use the most common 5000 English words. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-008 | linguistic_format | All the words that modify nouns should be verb participles. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-009 | linguistic_format | All sentences should use pattern of 'subjective + verb + objective + complement'. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-010 | linguistic_format | After the true subject appears at the beginning, use pronouns in the remaining sentences. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-011 | linguistic_format | Do not end each sentence with a noun. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-012 | linguistic_format | Write only interrogative sentences. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-013 | linguistic_format | Avoid first-person pronouns entirely. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-linguistic_format-014 | linguistic_format | Use no negation words. | hard | Responding with linguistic requirements, such as grammar, vocabulary, etc. | false | - Hint: you can use NLTK or spaCy python packages to verify the linguistic requirements, for example `import nltk` or `import spacy`. | 5 | 3 |
hard-prosody_phonetics_format-000 | prosody_phonetics_format | All words must contain exact three vowels. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-001 | prosody_phonetics_format | All words must contain a diphthong. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-002 | prosody_phonetics_format | Ensure every sentence has the pattern of increasing syllable count: the first has one syllable, the second two, and so on. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-003 | prosody_phonetics_format | Every sentence contains exactly 10 syllables. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-004 | prosody_phonetics_format | Write a three-line poem with syllable counts 5-7-5. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-005 | prosody_phonetics_format | Write a sentence where the first and last word rhyme. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-006 | prosody_phonetics_format | Write three sentences where the last word rhyme, | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-007 | prosody_phonetics_format | Write four sentences with alliteration. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-008 | prosody_phonetics_format | Create a sentence using three sound-imitating words. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-009 | prosody_phonetics_format | Ensure every word ends with a voiced consonant. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-prosody_phonetics_format-010 | prosody_phonetics_format | Write a limerick with AABBA rhyme and mark primary stress once per line. | hard | Responding with prosody or phonetics requirements on the words or sentences, such as rhythm, stress, prosody, rhyme scheme, vowels, consonant, etc. | false | - Hint: you can use Epitran python packages to verify the phonetic requirements, for example, `import epitran`.
- If the constraint implicitly requires the number of words, sentences, verify them as well. | 5 | 1 |
hard-inclusion_exclusion_types-000 | inclusion_exclusion_types | Use three different emojis in your response. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-001 | inclusion_exclusion_types | Come up with a 16-char password that includes at least one number and one special character. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-002 | inclusion_exclusion_types | Include a relevant citation with Chicago style (Author–Date). | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-003 | inclusion_exclusion_types | Include two relevant citations but not in IEEE style. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-004 | inclusion_exclusion_types | Include one URL in the response. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-005 | inclusion_exclusion_types | Include dates in your response and the dates should be in YYYYMMDD format. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-006 | inclusion_exclusion_types | Insert a single LaTeX inline equation enclosed in $...$. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-007 | inclusion_exclusion_types | Do not include any Markdown tables. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-inclusion_exclusion_types-008 | inclusion_exclusion_types | Do not include any tables of any kind, e.g., no Markdown tables, no CSV, no Latex tables, etc. | hard | Responding with or without a specific text type. | false | 5 | 4 | |
hard-consistency-000 | consistency | Respond with Markdown table and CSV table. | hard | Responding the same content but with different formats. | false | - The verification function should verify if all different formats exist and are valid, and then, verify if all formats present the same content. | 5 | 2 |
hard-consistency-001 | consistency | Respond with Markdown table and Latex table. | hard | Responding the same content but with different formats. | false | - The verification function should verify if all different formats exist and are valid, and then, verify if all formats present the same content. | 5 | 2 |
hard-consistency-002 | consistency | Respond with Markdown table, CVS, and Latex table. | hard | Responding the same content but with different formats. | false | - The verification function should verify if all different formats exist and are valid, and then, verify if all formats present the same content. | 5 | 2 |
hard-consistency-003 | consistency | List references in APA and Chicago; author names, titles, and years must match. | hard | Responding the same content but with different formats. | false | - The verification function should verify if all different formats exist and are valid, and then, verify if all formats present the same content. | 5 | 2 |
hard-consistency-004 | consistency | Provide the config with YAML and JSON. | hard | Responding the same content but with different formats. | false | - The verification function should verify if all different formats exist and are valid, and then, verify if all formats present the same content. | 5 | 2 |
hard-consistency-005 | consistency | Give your response in Markdown with ##, ### levels and in Latex with \section, \subsection. | hard | Responding the same content but with different formats. | false | - The verification function should verify if all different formats exist and are valid, and then, verify if all formats present the same content. | 5 | 2 |
hard-inclusion_exclusion_content-000 | inclusion_exclusion_content | Contain the word '[PLACEHOLDER]' for exact 5 times. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-001 | inclusion_exclusion_content | Use the two words '[PLACEHOLDER]' in your response, each at least 2 times. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-002 | inclusion_exclusion_content | Include the quote: '[PLACEHOLDER]' | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-003 | inclusion_exclusion_content | Make the two quotes in different paragraphs. The quotes '[PLACEHOLDER]'. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-004 | inclusion_exclusion_content | Include the following paragraph verbatim in your response. Paragraph:\n[PLACEHOLDER] | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-005 | inclusion_exclusion_content | Do not use the word '[PLACEHOLDER]' in your response. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-006 | inclusion_exclusion_content | Do not use these 5 forbidden words '[PLACEHOLDER]' in your response. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-007 | inclusion_exclusion_content | Do not include sentence '[PLACEHOLDER]' in your response. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-008 | inclusion_exclusion_content | Do not start each sentence with the phrase '[PLACEHOLDER]', but include it somewhere else at least once. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-009 | inclusion_exclusion_content | Do not start each sentence with the word '[PLACEHOLDER]', but include it somewhere else at least twice. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-010 | inclusion_exclusion_content | Answer with sentences that begin with the letter '[PLACEHOLDER]'. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-011 | inclusion_exclusion_content | Mention the word or phrase '[PLACEHOLDER]' only in the Level 1 markdown heading and nowhere else. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-inclusion_exclusion_content-012 | inclusion_exclusion_content | Use the synonym of '[PLACEHOLDER]' in the response and highlight them by ****. | hard | Containing or not containing a specific text content (e.g., letters, words, phrases, sentences, or even paragraphs), where the possible content is in [PLACEHOLDER] | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- If the [PLACEHOLDER] is supposed to contain multiple letters, words or phrases (e.g., the constraint specifically requires the placeholder to contain two/three words), expect they are separated by a comma or a semicolon when filling the placeholder. Make sur... | 5 | 5 |
hard-conversion-000 | conversion | Convert the given markdown table to Latex table. \nMarkdown table: \n[PLACEHOLDER] | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
hard-conversion-001 | conversion | Beautify the given JSON with 4-space indention. The given JSON object:\n[PLACEHOLDER] | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
hard-conversion-002 | conversion | Convert the input JSON to Python dict. The input JSON: \n[PLACEHOLDER] | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
hard-conversion-003 | conversion | Convert the given HTML to plaintext. The HTML: \n[PLACEHOLDER] | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
hard-conversion-004 | conversion | Convert the given CSV table to Latex table. The CSV: \n[PLACEHOLDER] | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
hard-conversion-005 | conversion | Replace [PLACEHOLDER] with its synonym. | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
hard-conversion-006 | conversion | Replace [PLACEHOLDER] with its antonym. | hard | Converting the structured (e.g., JSON, HTML, CSV) or semi-structured (e.g., Latex, Markdown) user text input, which is represented by the [PLACEHOLDER], to the target requirements. | true | - Hint: you can use NLTK WordNet to find synonyms/antonyms.
- Verify if the changing part is changed correctly and unchanging parts remain the same.
- Response may contain unstructured text, so verify if the target structure exists first. If not, return False. If so, extract it and verify if the content is consistent w... | 5 | 2 |
soft-tone-000 | tone | Respond in a formal tone. | soft | Responding with a specific and easily detected tone. | false | 5 | 3 |
End of preview. Expand in Data Studio
README.md exists but content is empty.
- Downloads last month
- 4