| { | |
| "task189-snli-neutral-to-contradiction-text-modification": "In this task, you're given a pair of sentences, sentence 1 and sentence 2, that neither agree with nor contradict each other. Your job is to alter sentence 2 so that the pair clearly contradict each other. Generated sentences must be short, with less than 15 words. New information can be introduced. Avoid using pronouns to confuse the subject of the sentence.", | |
| "task1534-daily-dialog-question-classification": "You are given a conversation between two people. 'Person1:' and 'Person2:' are used to separate their respective dialogues. If the conversation begins with a question, label it '1' otherwise '0'.", | |
| "task301-record-question-generation": "In this task, you will be shown a passage. You need to write a fill-in-the-gap question based on your understanding of the events that might be inferred from the passage. Your question should be answerable based on the passage and only have one correct answer. Show the gap in your question with a _ .", | |
| "task614-glucose-cause-event-detection": "In this task, you will be given a short story. One sentence from the story is chosen. Consider the events that happen before that sentence, or are likely to have happened before it. Does any of them directly cause it, or simply make it possible? You should write your answer in the form \" A >causes/enables> B\". Try to use phrases and sentences from the story to compose your answer when possible.", | |
| "task304-numeric-fused-head-resolution": "In this task, you will use your knowledge about language (and common sense) to determine what element the marked number refers to. The numbers are marked with two underlines around them, like: _ number _. There are several possible answers, you'll need to choose the proper one. Carefully read the given text, pay special attention to the marked number, think about what (unwritten) information the marked number holds inside, choose the most adequate word(s) from the optional answers. If none of them seems right to you, there's also an option for other. If your answer is \"REFERENCE\", also write the reference entity, otherwise write the implicit option name. Options to choose from are:\nREFERENCE: Some object which is being mentioned in the text before or after the target number. The reference answer has a higher priority than any other. If both Reference and another answer are possible, prioritize the Reference.\nYEAR: Describing a calendric year\nAGE: Describing someone's age\nCURRENCY: Reference to some monetary value e.g dollar, euro etc.\nPEOPLE: Describing a single/plural persons\nTIME: Describing a time of the day. Usually you can add the word o'clock after those numbers.\nOTHER: Some other option, which isn't listed here.", | |
| "task050-multirc-answerability": "You are given a sentence and a question in the input. If the information provided in the sentence is enough to answer the question, label \"Yes\", otherwise label \"No\". Do not use any facts other than those provided in the sentence while labeling \"Yes\" or \"No\". There are only two types of valid responses: Yes and No.", | |
| "task345-hybridqa-answer-generation": "In this task, you will be presented with a question, and you have to write the part-of-speech tag for each word in the question. Here is the Alphabetical list of part-of-speech tags used in this task: CC: Coordinating conjunction, CD: Cardinal number, DT: Determiner, EX: Existential there, FW: Foreign word, IN: Preposition or subordinating conjunction, JJ: Adjective, JJR: Adjective, comparative, JJS: Adjective, superlative, LS: List item marker, MD: Modal, NN: Noun, singular or mass, NNS: Noun, plural, NNP: Proper noun, singular, NNPS: Proper noun, plural, PDT: Predeterminer, POS: Possessive ending, PRP: Personal pronoun, PRP$: Possessive pronoun, RB: Adverb, RBR: Adverb, comparative, RBS: Adverb, superlative, RP: Particle, SYM: Symbol, TO: to, UH: Interjection, VB: Verb, base form, VBD: Verb, past tense, VBG: Verb, gerund or present participle, VBN: Verb, past participle, VBP: Verb, non-3rd person singular present, VBZ: Verb, 3rd person singular present, WDT: Wh-determiner, WP: Wh-pronoun, WP$: Possessive wh-pronoun, WRB: Wh-adverb", | |
| "task1598-nyc-long-text-generation": "The task is to write a full sentence or two using all of the information given. The sentence(s) will be a brief review of a restaurant. Use all of the information provided.", | |
| "task269-csrg-counterfactual-story-generation": "Given a premise, an initial context, an original ending, and a counterfactual context, the task is to generate a new story ending aligned with the counterfactual context and as close to the original ending as possible. Each instance consists of a five-sentence story. The premise is the first sentence of a story, and the second sentence, which is the initial context, provides more information about the story's context and the story's general plot. The original ending is the last three sentences of the story. Also, a counterfactual context is a slight modification to the initial context. You should write a new story ending that edits the original story ending as little as possible to regain coherence with the counterfactual context. To sum up, you should write the last three sentences of a story based on the premise(first sentence) and the counterfactual context(second sentence) of the story.", | |
| "task1186-nne-hrngo-classification": "You are given an original reference as well as a system generated reference. Your task is to judge the naturaleness of the system generated reference. If the utterance could have been produced by a native speaker output 1, else output 0.", | |
| "task030-winogrande-full-person": "In this task, based on the given context word, you are asked to create a pair of sentences each containing a blank (_). The sentence pair should look similar and should be about two different persons (PersonX and PersonY). Additionally, the two sentences must be different in terms of trigger words (e.g., \"sympathetic\" and \"stern\") which express contrasting attributes about the two persons. The answer to the first and the second sentence must be PersonX and PersonY, respectively. PersonX and PersonY should not be equally likely to fill the blank. For each sentence, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. Do not use the names of real people (e.g., Donald Trump, Putin, etc.) in your sentences. Avoid repeating the same style or phrase in twin sentences e.g., a contrasting sentence can always be created using simple negation i.e. by adding not, never, etc. Instead, try to increase diversity. Both twin sentences must contain at least 15 and at most 30 words. Twin sentences must have at least 70% overlapping words. You must utilize the given context word while writing the twin sentences. Each of the twin sentences must contain only one blank. Make sure that PersonX and PersonY have the same gender. In each sentence, PersonX and PersonY should be used only ONCE and PersonX should appear earlier than PersonY.", | |
| "task905-hate-speech-offensive-classification": "You are given a text of the tweet and a corresponding label whether this tweet is 'Offensive', 'Hate Speech' or 'Neither'. Your job is to identify if the label is correct. Generate label 'true' if it's correct, 'false' otherwise.", | |
| "task1359-numer-sense-answer-generation": "Given a sentence, fill out the missing word with a 'no' or a number (between zero and ten). You should write the numbers with english alphabet, like: four instead of 4.", | |
| "task502-scruples-anecdotes-whoiswrong-verification": "Given a post that is a real-life anecdote of a complex ethical situation and a question asks if AUTHOR, NOBODY, EVERYBODY, or OTHER is wrong in the situation. Answer the question and classify your answers into yes or no.", | |
| "task080-piqa-answer-generation": "In this task the focus is on physical knowledge about the world. Given the provided goal task in the input, describe a process that would lead to the asked outcome. This process often involves physical motions with objects, such as moving them, arranging them in a certain way, mixing them, shaking them, etc.", | |
| "task1368-healthfact-sentence-generation": "Given a paragraph, generate a claim that is supported by the given paragraph. 1) The claim must contain information from within the paragraph. 2) A sentence within the paragraph can be used as a claim. 3) The claim should not have contradictions within the paragraph. 4) The claim should be at most one sentence long.", | |
| "task469-mrqa-answer-generation": "Given a question and a context passage, generate the answer having the word or phrase from the context passage. Here, the answer should be a shortest continous span from the passage.", | |
| "task430-senteval-subject-count": "In this task you are given a sentence. You must judge whether subject of the main clause is singular or plural. Label the instances as \"Singular\" or \"Plural\" based on your judgment.", | |
| "task280-stereoset-classification-stereotype-type": "In this task, you are given a short passage that conveys stereotype or anti-stereotype about a specific target. A stereotype is an over-generalized belief about a particular group of people. An anti-stereotype is an idea that goes against a common stereotype. You are expected to classify the passage into four types of stereotype or anti-stereotype: gender, profession, race, and religion.", | |
| "task186-snli-contradiction-to-entailment-text-modification": "In this task, you're given a pair of sentences, sentence 1 and sentence 2, that contradict each other. Your job is to alter sentence 2 so that the pair agree with each other. Generated sentences must be short, with less than 15 words. New information can be introduced. Avoid using pronouns to confuse the subject of the sentence.", | |
| "task268-casehold-legal-answer-generation": "In this task, you will be shown a prompt from a judicial decision and multiple holding statements derived from citations following text in a legal decision. Holdings represent the governing legal rule when the law is applied to a particular set of facts. There are five answer choices for each citing text. The correct answer is the holding statement that corresponds to the citing text. The four incorrect answers are other holding statements. You should find the correct option. There is a <HOLDING> token in the position of the citing text prompt where the holding statement was extracted.", | |
| "task1206-atomic-classification-isbefore": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether the Head happens before the Tail or not. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task024-cosmosqa-answer-generation": "Craft one correct answer to the question given in input. To make it more interesting, try to use non-stereotypical language if possible. Make sure your correct answer is reasonably long, consistent with the context, and requires common sense (instead of explicit extraction from the context.) In your answer, use as few words as possible from the given context. Use a response that is uncommon/non-stereotypical, so that it is less predictable. To be less repetitive, please vary your language for each question.", | |
| "task579-socialiqa-classification": "In this task, you're given a context, a question, three options, and an answer. Your task is to classify whether the given answer is correct or not by providing 'Yes' or 'No', based on the context with commonsense reasoning about social situations.", | |
| "task1552-scitail-question-generation": "Given scientific fact as input, generate the question from this fact such that it can be answered from the input.", | |
| "task184-snli-entailment-to-neutral-text-modification": "In this task, you're given a pair of sentences, sentence 1 and sentence 2, that agree with each other. Your job is to alter sentence 2 so that the pair neither agree nor contradict each other. Generated sentences must be short, with less than 15 words. New information can be introduced. Avoid using pronouns to confuse the subject of the sentence.", | |
| "task1411-dart-subject-identification": "This task is about identifying the subject of a given sentence. The subject of a sentence is the person, place, or thing that is performing the action of the sentence. The subject represents what or whom the sentence is about. Each sentence in this task is guaranteed to contain a single subject. The input is a sentence and the output is the identified subject of that sentence.", | |
| "task924-event2mind-word-generation": "You are provided with an \"Event\", \"Intent\" and \"XEmotion\" (PersonX's reactions for the given \"Event\"). Indicate PersonY's reaction (person feels) at the end of this event. Provide one reaction for PersonY. If there's nothing that can be implied, respond as None", | |
| "task1412-web-questions-question-answering": "A question is presented to you in this task, and your job is to write a potentially correct answer.", | |
| "task678-ollie-actual-relationship-answer-generation": "From the given sentence, extract the phrase (often noun or verb) that carries the given relationship. The generated phrase should be present in the sentence.", | |
| "task194-duorc-answer-generation": "In this task, you will be shown an extract from a movie plot and a question. You need to provide the correct answer for it. Short answers containing words that are present in the passage are preferred.", | |
| "task346-hybridqa-classification": "In this task, you will be presented with a question, a word, and a POS tag. You have to determine whether the part-of-speech tag of the given word in the question is equal to the given POS tag or not. Give your answer with True or False. Here is the Alphabetical list of part-of-speech tags used in this task: CC: Coordinating conjunction, CD: Cardinal number, DT: Determiner, EX: Existential there, FW: Foreign word, IN: Preposition or subordinating conjunction, JJ: Adjective, JJR: Adjective, comparative, JJS: Adjective, superlative, LS: List item marker, MD: Modal, NN: Noun, singular or mass, NNS: Noun, plural, NNP: Proper noun, singular, NNPS: Proper noun, plural, PDT: Predeterminer, POS: Possessive ending, PRP: Personal pronoun, PRP$: Possessive pronoun, RB: Adverb, RBR: Adverb, comparative, RBS: Adverb, superlative, RP: Particle, SYM: Symbol, TO: to, UH: Interjection, VB: Verb, base form, VBD: Verb, past tense, VBG: Verb, gerund or present participle, VBN: Verb, past participle, VBP: Verb, non-3rd person singular present, VBZ: Verb, 3rd person singular present, WDT: Wh-determiner, WP: Wh-pronoun, WP$: Possessive wh-pronoun, WRB: Wh-adverb", | |
| "task844-financial-phrasebank-classification": "Given a piece of financial news and its polarity, classify it into 'true' if the polarity is correct and classify into 'false' if the polarity is incorrect. Output must be 'true' or 'false'.", | |
| "task163-count-words-ending-with-letter": "In this task, you need to count the number of words in a sentence that end with the given letter. Answer with numbers and not words.", | |
| "task151-tomqa-find-location-easy-clean": "Given a story, answer the question about the story. The question is the last sentence in the input. The story has one of the three following scenarios: (1) when the individual's belief matches reality, (2) when the individual's belief does not match reality, (3) is when an individual has a false belief about another individual's beliefs. The question will ask about the location of an object in the story with respect to either none or one of the three scenarios.", | |
| "task616-cola-classification": "You're given a sentence and your task is to classify whether the sentence is acceptable or not. Any sentence which is grammatically correct, has a naturalistic text, is written by a native speaker and which minimizes superfluous content is acceptable, otherwise unacceptable. If the sentence is acceptable then write \"acceptable\", otherwise \"unacceptable\".", | |
| "task1602-webquestion-question-genreation": "You are given a concept, and a list of answers. You should generate a question about the concept that leads to the given answer(s).", | |
| "task075-squad1.1-answer-generation": "This task is about writing a correct answer for the reading comprehension task. Based on the information provided in a given passage, you should identify the shortest continuous text span from the passage that serves as an answer to the given question. Avoid answers that are incorrect or provides incomplete justification for the question.", | |
| "task1211-atomic-classification-hassubevent": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether the Head includes an event or an action in the Tail or not. This happens when the Tail denotes a step within the larger head event. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task072-abductivenli-answer-generation": "In this task, you're given the middle and ending of a three-part story. Your job is to complete the short story by writing a probable beginning of the story. Generated sentences must be short, have fewer than 10 words, and be simple as if narrating to a child. Avoid using any irrelevant extra information when creating the beginning, and use names (e.g. Jack, Barbara,...) instead of pronouns (e.g. he / she) wherever possible.", | |
| "task350-winomt-classification-gender-identifiability-pro": "In this task, you are given a sentence and a profession. The sentence mentions two professions: one's gender is identifiable using the coreference link with gendered pronouns and the other's gender is unidentifiable. You are expected to return whether the given profession's gender is identifiable or unidentifiable.", | |
| "task333-hateeval-classification-hate-en": "In this task, you are given a post in English from online platforms. You are expected to identify whether the post is hateful against immigrants and women. A hateful post expresses hate or encourages violence towards a person or a group. If a post is hateful but not towards immigrants and women, it should be labeled as non-hateful. Answer \"hateful\" or \"Non-hateful\". Note that the URLs in the text have been replaced with [Link].", | |
| "task1293-kilt-tasks-hotpotqa-question-answering": "In this task, you are given a question. You have to answer the question based on your information.", | |
| "task683-online-privacy-policy-text-purpose-answer-generation": "Given a part of privacy policy text, identify the purpose for which the user information is collected/used. The purpose should be given inside the policy text, answer as 'Not Specified' otherwise", | |
| "task508-scruples-dilemmas-more-ethical-isidentifiable": "In this task, you are given a pair of action statements. Your task is to check if you can conclusively identify which statement is less ethical or not. Answer: 'no' if one of the statements is less ethical, and 'yes' otherwise.", | |
| "task570-recipe-nlg-ner-generation": "In this task, you have to generate the named entities (NER) given its ingredients of the recipe. Named entities are the names of the items without their quantity.", | |
| "task1727-wiqa-what-is-the-effect": "In this task you will be given a process, and a question. The process contains a sequence of steps that happen in order. The question asks about the effect of a certain event on another event. If the first event has a positive effect on the second event, answer with \"for\", if it has a negative effect, answer with \"against\". If there's no causal relationship between the two, answer with \"none\".", | |
| "task001-quoref-question-generation": "In this task, you're given passages that contain mentions of names of people, places, or things. Some of these mentions refer to the same person, place, or thing. Your job is to write questions that evaluate one's understanding of such references. Good questions are expected to link pronouns (she, her, him, his, their, etc.) or other mentions to people, places, or things to which they may refer. Do not ask questions that can be answered correctly without understanding the paragraph or having multiple answers. Avoid questions that do not link phrases referring to the same entity. For each of your questions, the answer should be one or more phrases in the paragraph, and it should be unambiguous.", | |
| "task1532-daily-dialog-emotion-classification": "You are given a conversation between two people. 'Person1:' and 'Person2:' are used to separate their respective dialogues. Your task is to label each of the continuous conversations done by one speaker with the emotion detected in that particular part. Your labels can be one among the following: 'No emotion', 'surprise', 'happiness', 'sadness', 'anger', 'disgust'.", | |
| "task349-squad2.0-answerable-unanswerable-question-classification": "In this task, you are given a passage and a question regarding that passage. You must determine whether or not the question is answerable from the given passage. If a question is answerable, output should be 'True', otherwise 'False'. You must not consider any other information that is not provided in the passage while labelling True or False.", | |
| "task1533-daily-dialog-formal-classification": "You are given a conversation between two people. 'Person1:' and 'Person2:' are used to separate their respective dialogues. You are required to assign a label 'formal' if there is an absence of emotion and a presence of questions anywhere within the conversation. If such pattern is not found assign the label 'informal'.", | |
| "task351-winomt-classification-gender-identifiability-anti": "In this task, you are given a sentence and a profession. The sentence mentions two professions: one's gender is identifiable using the gendered pronouns in the text and the other's gender is unidentifiable. You are expected to return whether the given profession's gender is identifiable or unidentifiable.", | |
| "task212-logic2text-classification": "In this task, you are given commands (in terms of logical operations) to select relevant rows from the given table. Your job is to classify the command into one of these seven categories: (1) majority, (2) unique, (3) superlative, (4) count, (5) comparative, (6) aggregation, and (7) ordinal. \n Here are the defications of each category: \n 1. majority: Describing the majority values (most or all) over one column, with the scope of all table rows or a subset of rows \n 2. unique: Describing one unique row, regarding one column, with the scope of all table rows or a subset of rows \n 3. Superlative: Describing the maximum or minimum value in a column, with the scope of all table rows or a subset of rows \n 4. Ordinal: Describing the n-th maximum or minimum value in a column, with the scope of all table rows or a subset of rows \n 5. Comparative: Comparing two rows in the table, regarding their values in one column \n 6. Count: counting some rows in the table based on the values in one column, with the scope of all table rows or a subset of rows \n 7. Aggregation: Describing the sum or average value over a column, with the scope of all table rows or a subset of rows. \n Here are the definitions of logical operators for understanding of command: \n 1. count: returns the number of rows in the view. \n 2. only: returns whether there is exactly one row in the view. \n 3. hop: returns the value under the header column of the row. \n 4. and: returns the boolean operation result of two arguments. \n 5. max/min/avg/sum: returns the max/min/average/sum of the values under the header column. \n 6. nth_max/nth_min: returns the n-th max/n-th min of the values under the header column. \n 7. argmax/argmin: returns the row with the max/min value in header column. \n 8. nth_argmax/nth_argmin: returns the row with the n-th max/min value in header column. \n 9. eq/not_eq: returns if the two arguments are equal. \n 10. round_eq: returns if the two arguments are roughly equal under certain tolerance. \n 11. greater/less: returns if the first argument is greater/less than the second argument. \n 12. diff: returns the difference between two arguments. \n 13. filter_eq/ filter_not_eq: returns the subview whose values under the header column is equal/not equal to the third argument. \n 14. filter_greater/filter_less: returns the subview whose values under the header column is greater/less than the third argument. \n 15. filter_greater_eq /filter_less_eq: returns the subview whose values under the header column is greater/less or equal than the third argument. \n 16. filter_all: returns the view itself for the case of describing the whole table \n 17. all_eq/not_eq: returns whether all the values under the header column are equal/not equal to the third argument. \n 18. all_greater/less: returns whether all the values under the header column are greater/less than the third argument. \n 19. all_greater_eq/less_eq: returns whether all the values under the header column are greater/less or equal to the third argument. \n 20. most_eq/not_eq: returns whether most of the values under the header column are equal/not equal to the third argument. \n 21. most_greater/less: returns whether most of the values under the header column are greater/less than the third argument. \n 22. most_greater_eq/less_eq: returns whether most of the values under the header column are greater/less or equal to the third argument.", | |
| "task213-rocstories-correct-ending-classification": "In this task, you're given the title of a five-sentence story, the first four sentences, and two options for the fifth sentence as a and b. Your job is to pick the sentence option that seamlessly connects with the rest of the story, indicating your choice as 'a' or 'b'. If both sentences are plausible, pick the one that makes more sense.", | |
| "task401-numeric-fused-head-reference": "In this task, you will use your knowledge about language (and common sense) to determine what element the marked number refers to. The numbers are marked with two underlines around them, like: _ number _. Your answer should be chosen from the given text, and should not contain other words.", | |
| "task1326-qa-zre-question-generation-from-answer": "Generate a question which can yield the answer mentioned in the input. Generated question must be answered by the answer provided in input, without using any extra knowledge.", | |
| "task1599-smcalflow-classification": "In this task, you are given an utterance, which is a part of a conversation between a user and an agent. Your job is to detect the speaker. The user usually commands the agent to schedule events, retrieve event information, and check the schedule. While the agent's response is an answer to the user's questions or follow-up questions on the user's command. Answer with \"User\" or \"Agent\".", | |
| "task218-rocstories-swap-order-answer-generation": "In this task, you're given the title of a story consisting of five sentences, numbered 1 through 5. Your job is to determine which two sentences need to be swapped sentences in order to make a story that makes complete sense and is befittingly titled. Indicate your answer using the numbers of the two sentences in order, such as '34' or '25'. The first digit refers to the sentence which should come first in the story.", | |
| "task379-agnews-topic-classification": "In this task, you are given a news article. Your task is to classify the article to one out of the four topics 'World', 'Sports', 'Business', 'Sci/Tech' if the article's main topic is relevant to the world, sports, business, and science/technology, correspondingly. If you are not sure about the topic, choose the closest option. Note that URLs in the text have been replaced with [Link].", | |
| "task184-break-generate-question": "In this task, you will be given a set of steps that are required to answer a specific question. Your job is to generate that question. Each given step refers to either an entity (known or unknown), a propery of an entity or a query operation (count, group, union, etc.) Knowing those operations and how they appear in the input may help you generate more accurate questions.\n Select: A select step is used to return a set of objects. There are no references to previous steps in a select step. template: Return [attributes]\n Filter: A filter step is used to return results from a previous step to which a certain condition applies. template: Return [#step] [condition]\n Project: A project step should return certain attributes of the results of a previous step. template: Return [attributes] of [#step]\n Aggregate: An aggregate step returns an aggregator function applied on a step's result. template: Return the [aggregator] of [#step].\n Group: A group step is an aggregator applied on attributes. template: Return the [aggregator] of [#step] for each [attribute]\n Superlative: A superlative step is used to return the result with a highest/lowest attribute among other results. template: Return [#step1] [where] [#step2] [is] [highest / lowest]\n Comparative: A comparative step is used when we need to compare an attribute with a number to filter results. template: Return [#step1] [where] [#step2] [comparator] [number] \n Union: A union step is used to return results of two steps together. template: Return [#step1] [or / ,] [#step2]\n Intersection: An intersection step returns the result that two steps have in common. template: Return [attribute] of both [#step1] and [#step2]\n Discard: A discard step returns result of a step and excludes result of another step from it. template: Return [#step1] besides [#step2]\n Sort: A sort returns result of another step in a specific order. template: Return [#step1] [ordered / sorted by] [#step2]\n Is true: An is true step checks a condition on another result and returns a true or false. template: Return [is / if] [condition]\n Arithmetic: An arithmatic step operates an arithmatic operation on one or more steps. template: Return the [arithmetic op.] of [#step1] [and] [#step2].", | |
| "task470-mrqa-question-generation": "Given a context passage, generate a question from the passage such that its answer is shortest continous span from the passage.", | |
| "task923-event2mind-classifier": "You are provided with an \"Event\" and it's \"Intent\" related to PersonX. Determine the sentiment value of the given input as either \"Positive\", \"Negative\", and \"Unknown\".", | |
| "task221-rocstories-two-choice-classification": "In this task, you're given the title and three arbitrary sentences out of a five-sentence story. You are also given three additional sentence options, a, b, and c, that may or may not belong to the story. Your job is to pick the two options that seamlessly connect with the rest of the story; note that the selected choices may fit into the story at any point. Indicate your selected choices in order, i.e., as 'ab'/'bc'/'ac' and not as 'ba'/'ca'/'cb'. If options are equally plausible, pick the ones that make more sense.", | |
| "task192-hotpotqa-sentence-generation": "In this task, you are given a context paragraph, a question based on that and corresponding answer of a question. Your task is to generate supporting fact/knowledge from context paragraph which helps in answering a given question.", | |
| "task1391-winogrande-easy-answer-generation": "In this task, you are given a question containing a blank (_) and two options. You should pick the best option to answer the question. Please answer with \"A\" or \"B\".", | |
| "task241-tweetqa-classification": "In this task, you are given a context tweet, a question and corresponding answer of given question. Your task is to classify this question-answer pair into two categories: (1) \"yes\" if the given answer is right for question, and (2) \"no\" if the given answer is wrong for question.", | |
| "task638-multi-woz-classification": "You are shown a conversation between a user and system. Identify who has spoken the indicated sentence based on the conversation.", | |
| "task114-is-the-given-word-longest": "In this task, you need to answer 'Yes' if the given word is the longest word (in terms of number of letters) in the given sentence, else answer 'No'. Note that there could be multiple longest words in a sentence as they can have the same length that is the largest across all words in that sentence.", | |
| "task161-count-words-containing-letter": "In this task, you need to count the number of words in a sentence that contain the given letter", | |
| "task065-timetravel-consistent-sentence-classification": "In this task, you are given a short story consisting of exactly 5 sentences where the second sentence is missing. You are given two options and you need to select the one that best connects the first sentence with the rest of the story. Indicate your answer by 'Option 1' if the first option is correct, otherwise 'Option 2'. The incorrect option will change the subsequent storyline, so that at least one of the three subsequent sentences is no longer consistent with the story.", | |
| "task031-winogrande-question-generation-object": "You are asked to create a question containing a blank (_), based on the given context word. Your question must contain two related but different objects; for example \"trophy\" and \"suitcase\". The expected answer to your question must be one of the objects present in the sentence. The expected answer must not be associated with any specific word in the question; instead it should depend on the context present in the question. The expected answer should not be equally likely to fill the blank. For your question, there should be a agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. Do not use animals or proper nouns (e.g., New York, Macbook, Jeff Bezos, McDonald's, ...) as your objects. Avoid repeating the same style, pattern or phrases in each question, try to increase diversity by varying sentence structure, blank placement etc. Your question must contain at least 15 and at most 30 words. You must utilize the given context word while writing the question. Your question must contain only one blank. Make sure that Object X and Y have the same number e.g. when ObjectX is singular, Object Y must be singular, too. The two objects (Object X & Object Y) should be used ONCE in your question. Here is a list of attributes and associated pair of contrastive words which may be used to create a valid question using the objects. You can use either of the contrastive words, but not both. You should think about more such attributes and associated words and use them in your question.\n| Attribute | triggerword | contrastive triggerword | \n| age | old | new | \n| altitude | low | high |\n| area | small | vast | \n| brightness | dark | light | \n| clarity | obscure | clear | \n| cleanness | dirty | clean | \n| complexity | simple | complex | \n| cost | cheap | expensive |\n| density | sparse | dense |\n| depth | shallow | deep |\n| distance | near | far | \n| electric conductivity | low | high |\n| flexibility | rigid | flexible |\n| granularity | fine | coarse | \n| hardness | soft | hard | \n| length | short | long | \n| magnitude | small | large | \n| mass | small | large | \n| odor | weak | strong | \n| pressure | low | high | \n| resistance | low | high | \n| shape | round | sharp |\n| shape | flat | spiky |\n| size | small | large | \n| sound | quiet | loud | \n| sound pitch | low | high | \n| speed | slow | fast | \n| stability | unstable | stable | \n| strength | weak | strong | \n| temperature | low | high | \n| texture | smooth | rough | \n| thermal conductivity | low | high |\n| thickness | thin | thick | \n| volume | small | large |\n| weight | light | heavy | \n| width | narrow | wide | \n| location | in | out |\n| location | up | down |\n| location | above | below |\n| location | on | off |\n| location | to | from |", | |
| "task199-mnli-classification": "In this task, you're given a pair of sentences, sentence 1 and sentence 2. Your job is to determine if the two sentences clearly agree/disagree with each other, or if this can't be determined. Indicate your answer as yes or no respectively.", | |
| "task323-jigsaw-classification-sexually-explicit": "In this task, you are given a public comment from online platforms. You are expected to classify the comment into two classes: sexual-explicit and non-sexual-explicit. A comment is considered sexual-explicit if it explicitly portrays sexual matters.", | |
| "task1555-scitail-answer-generation": "In this task, you are given a sentence and a question, you would be asked to create the answer which is contained in the sentence provided.", | |
| "task501-scruples-anecdotes-post-type-verification": "Given a post that is a real-life anecdote of a complex ethical situation and an associated claim about its type, verify if the claim is true or not. The claim asks if the posts are historical or hypothetical. The posts are \"HISTORICAL\" when the author has already done something and they are \"HYPOTHETICAL\" when the author is considering doing something in the future. Say 'yes' if the claim is true otherwise say 'no'", | |
| "task309-race-answer-generation": "In this task, you're given an article, a question which often contains a blank and four options (associated with \"A\", \"B\", \"C\", \"D\"). Your task is to find the correct answer (from the given options) for the question from the given article and return one of the options from \"A\", \"B\", \"C\", and \"D\". Do not generate anything else apart from one of the following characters: \"A\", \"B\", \"C\", \"D\". There is only one correct answer for each question.", | |
| "task203-mnli-sentence-generation": "In this task, you're given a statement, the genre to which that statement belongs, and a label indicating if the statement should be agreed with (entailment), disagreed with (contradiction), or neither (neutral). Your job is to write a sentence that describes the genre that follows the tone with respect to the statement, as indicated by the label. If sentence X agrees with sentence Y, the can be concluded from one another. If sentence X disagrees with sentence Y, they can not be correct at the same time. The sentence must also belong to the genre specified.", | |
| "task310-race-classification": "In this task, you're given an article, a question which often contains a blank, four options (associated with \"A\", \"B\", \"C\", \"D\") and the answer to that question. Your task is to classify whether the given answer is correct or not by providing \"Yes\" or \"No\", based on the article.", | |
| "task1658-billsum-summarization": "In this task, you are given text for US Congressional and California state bills, your task is to generate a summary for this bill.", | |
| "task028-drop-answer-generation": "This task involves creating answers to complex questions, from a given passage. Answering these questions, typically involve understanding multiple sentences. Make sure that your answer has the same type as the \"answer type\" mentioned in input. The provided \"answer type\" can be of any of the following types: \"span\", \"date\", \"number\". A \"span\" answer is a continuous phrase taken directly from the passage or question. You can directly copy-paste the text from the passage or the question for span type answers. If you find multiple spans, please add them all as a comma separated list. Please restrict each span to five words. A \"number\" type answer can include a digit specifying an actual value. For \"date\" type answers, use DD MM YYYY format e.g. 11 Jan 1992. If full date is not available in the passage you can write partial date such as 1992 or Jan 1992. If you find multiple spans, please add them all as a comma separated list. Please restrict each span to five words.", | |
| "task029-winogrande-full-object": "In this task, based on the given context word, you are asked to create a pair of sentences each containing a blank (_) and their corresponding answer. The sentence pair should look similar, and should be about two related but different objects; for example \"trophy\" and \"suitcase\". Additionally, the two sentences must be different in terms of trigger words (e.g., \"small\" and \"big\") which express contrasting properties about the two objects. The answer must not be associated with the trigger word; instead, it should depend on the context present in twin sentences. Also, the answers should not be ambiguous. For each sentence, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. Do not use animals and proper nouns (e.g., New York, Macbook, Jeff Bezos, McDonald's, ...) as your objects. Avoid repeating the same style or phrase in twin sentences. E.g., a contrasting sentence can always be created using simple negation i.e. by adding not, never, etc. Instead, try to increase diversity. Both twin sentences must contain at least 15 and at most 30 words. Twin sentences must have at least 70% overlapping words. You must utilize the given context word while writing the twin sentences. Each of the twin sentences must contain only one blank. Make sure that ObjectX and Y have the same number e.g. when ObjectX is singular, ObjectY must be singular, too. The two objects (ObjectX & ObjectY) should be used ONCE in each sentence. Here is a list of contrastive words that may be used as trigger words. You should create more such trigger words and use them in your twin sentences.\n| Attribute | triggerword | contrastive triggerword | \n| age | old | new | \n| altitude | low | high |\n| area | small | vast | \n| brightness | dark | light | \n| clarity | obscure | clear | \n| cleanness | dirty | clean | \n| complexity | simple | complex | \n| cost | cheap | expensive |\n| density | sparse | dense |\n| depth | shallow | deep |\n| distance | near | far | \n| electric conductivity | low | high |\n| flexibility | rigid | flexible |\n| granularity | fine | coarse | \n| hardness | soft | hard | \n| length | short | long | \n| magnitude | small | large | \n| mass | small | large | \n| odor | weak | strong | \n| pressure | low | high | \n| resistance | low | high | \n| shape | round | sharp |\n| shape | flat | spiky |\n| size | small | large | \n| sound | quiet | loud | \n| sound pitch | low | high | \n| speed | slow | fast | \n| stability | unstable | stable | \n| strength | weak | strong | \n| temperature | low | high | \n| texture | smooth | rough | \n| thermal conductivity | low | high |\n| thickness | thin | thick | \n| volume | small | large |\n| weight | light | heavy | \n| width | narrow | wide | \n| location | in | out |\n| location | up | down |\n| location | above | below |\n| location | on | off |\n| location | to | from |", | |
| "task027-drop-answer-type-generation": "This task involves annotating the answer type to a given question that involve some kind of complex reasoning (including numerical reasoning). Note that the questions require looking at more than one part of the passage to answer. There are 3 possible answer types (i) spans, (ii) numbers and (iii) dates. If the answer can be found in the passage, label it as \"span\". If the answer is a number, label as \"number\". Similarly, label \"date\" if you think the answer to the given question is a date.", | |
| "task400-paws-paraphrase-classification": "In this task you are given a sentence pair from wikipedia that have high lexical overlap. If the sentences have the same meaning and are just paraphrases of each other label them as \"Paraphrase\" , if not label them as \"Not paraphrase\". The two sentences are seperated by a new line.", | |
| "task673-google-wellformed-query-classification": "Read the given query and classify it as a 'Good' or 'Bad' query depending on how well the query is formed, 'Bad' being the expected output for a not so well formed query and 'Good' being the expected output for a well formed query. A query may be wrong based on common sense or general facts, but if it is well formed, you should answer with Good.", | |
| "task328-jigsaw-classification-insult": "In this task, you are given a public comment from online platforms. You are expected to classify the comment into two classes: insult and non-insult. Insult is any lanugage or act that is disrespectful or scornfully abusive.", | |
| "task1422-mathqa-physics": "In this task, you need to answer the given multiple-choice question on the physics. Classify your answers into 'a', 'b', 'c', 'd', and 'e'.", | |
| "task846-pubmedqa-classification": "Given a passage with a question and an answer to that question, classify if the answer actually answers the question into 'yes' or 'no'. Output 'yes' if the answer answers the question and output 'no' if the answer does not answer the question.", | |
| "task403-creak-commonsense-inference": "In this task you are given a statement and an explanation giving you further knowledge about an entity in the statement. You must judge whether the statement is true or false based on the explanation. Label an instance as \"True\" if the explanation confirms the statement or doesn't disprove it. Label an instance as \"False\" if the explanation disproves the statement. The statement and explanation are separated by a newline character.", | |
| "task290-tellmewhy-question-answerability": "In this task you are given a story and a question regarding that story. You must judge whether the question is answerable based on the info given to you. Label the instances as \"Answerable\" or \"Not Answerable\" based on your judgment. the story and the question are separated by a new line character.", | |
| "task1595-event2mind-text-generation-1": "The task is to generate text based of an event. For the given event, there will be a person who does the action in the event (usually PersonX). Your task is to write that persons emotional reaction to the event, i.e. how that person feels after doing the action.", | |
| "task1541-agnews-classification": "In this task, you're given a short article. Your job is to classify the article based on its category. Use the following classification labels, 0. World, 1. Sports, 2. Business, 3. Science or Technical. Label the text \"0\" if it contains information related to world. Label the text \"1\" if it contains information related to sports. Label the text \"2\" if it contains information related business. Label the text \"3\" if it contains science or technical related information.", | |
| "task848-pubmedqa-classification": "Given a passage classify if the passage has a definite objective/aim/goal or not. Output '1' if the passage has a defininte objective/aim/goal and output '0' if the passage does not have a definite objective/aim/goal.", | |
| "task569-recipe-nlg-text-generation": "In this task, you have to generate the title of the recipe given its required ingredients and directions.", | |
| "task582-naturalquestion-answer-generation": "In this task, You are given an open-domain question that can be answered based on factual information. Your task is to provide \\*short\\* answer (in a few words only) for the given question. The short answer can be one or more entities or it can also be boolean \\*yes\\* or \\*no\\*.", | |
| "task168-strategyqa-question-decomposition": "In the following task, you are given a yes/no question, its answer, and additional information which includes a description of a key term in the question and several facts. Your task is to \"decompose the question\"i.e. write the steps required to construct the given answer where each step is a question that can be answered using Wikipedia articles. To decompose the question, think of a decomposition strategy you could apply to answer the question through a search against Wikipedia. The decomposition strategy might not be obvious for some questions. For this reason, you are provided with additional information. You are not required to write decompositions that exactly match the provided facts, they are only meant as guidance. If the facts are too noisy, think about how you would have used Wikipedia to answer this question to create the decomposition. \"not_definitive_answer\" is true if there is no clear answer i.e. the question is ambiguous or a subjective opinion.\"incorrect_answer\" is true if the answer to the question is not correct. If the answer is incorrect, write a decomposition that evaluates the correct answer. Each question should be easy to answer based on simple facts from Wikipedia pages, or a simple check using the answers to preceding questions. To ensure this, we also ask you to enter a potential Wikipedia article that would contain the answer to each question. If you can not think of a Wikipedia page that can answer your question, please think of a different strategy. You do not need to enter a Wikipedia page for steps that don't add new information and only contain simple checks. The answer to the final question is the same as the answer to the original question. Every question's answer has an impact on the final answer and there should be no redundant steps in the decomposition.", | |
| "task413-mickey-en-sentence-perturbation-generation": "Given a sentence, generate a new sentence by performing small changes on the sentence. Here, make sure that the changes are semantically related and syntactically similar to the input. And the generated sentence should have high commonsense plausibility, that is to have reasonable probability of it being true.", | |
| "task390-torque-text-span-selection": "In this task, you will be given a passage, a question and a list of single-word events identified from the passage. The goal is to point out all events from the given event list that can answer the question. If a question is unanswerable or none of the events answer the question, output should be empty.", | |
| "task382-hybridqa-answer-generation": "In this task, you will be presented with a question about part-of-speech tag of a word in the question. You should write the required POS tag answering the question. Here is the Alphabetical list of part-of-speech tags used in this task: CC: Coordinating conjunction, CD: Cardinal number, DT: Determiner, EX: Existential there, FW: Foreign word, IN: Preposition or subordinating conjunction, JJ: Adjective, JJR: Adjective, comparative, JJS: Adjective, superlative, LS: List item marker, MD: Modal, NN: Noun, singular or mass, NNS: Noun, plural, NNP: Proper noun, singular, NNPS: Proper noun, plural, PDT: Predeterminer, POS: Possessive ending, PRP: Personal pronoun, PRP$: Possessive pronoun, RB: Adverb, RBR: Adverb, comparative, RBS: Adverb, superlative, RP: Particle, SYM: Symbol, TO: to, UH: Interjection, VB: Verb, base form, VBD: Verb, past tense, VBG: Verb, gerund or present participle, VBN: Verb, past participle, VBP: Verb, non-3rd person singular present, VBZ: Verb, 3rd person singular present, WDT: Wh-determiner, WP: Wh-pronoun, WP$: Possessive wh-pronoun, WRB: Wh-adverb", | |
| "task279-stereoset-classification-stereotype": "In this task, you are given a short passage that may convey stereotype, anti-stereotype, or is unrelated. A stereotype is an over-generalized belief about a particular group of people. An anti-stereotype is an idea that goes against a common stereotype. The passage is unrelated if it does not convey a stereotype or anti-stereotype. You are expected to classify the passage into three classes: stereotype, anti-stereotype, and unrelated.", | |
| "task191-hotpotqa-question-generation": "In this task, you are given a set of context paragraphs, some supporting facts and an answer of a question. Your task is to generate question for given answer based on set of context paragraphs, supporting facts and an answer.", | |
| "task1369-healthfact-sentence-generation": "Generate an explanation for the given claim using the provided supporting material from the paragraph. Please consider the following points while generating an output. 1) The claim will always have supporting proof in the paragraph, and the paragraph will have a clear point of view supporting the claim. 2) There can be instances where direct correlation to the claim is not available. Make use of the sentences in the paragraph to provide a corresponding output supporting the claim. 3) The explanation should be at least one and at most twelve sentences long.", | |
| "task061-ropes-answer-generation": "You are given a background paragraph that describes one or more causal or qualitative relationships such as a relationship in economics or a scientific law and a story that makes use of the concepts or the relationship described in the provided paragraph. You are also given a question about the story that requires an understanding of the relationship described in the background paragraph and the story. You need to come up with an answer to the given question; the answer will be a span from either the question or the story. In order to correctly answer the given question, you need to understand the relationship mentioned in the background paragraph and should be able to use it to understand that in the story. Your answer can not consist of any word that is not mentioned in any of these: the background paragraph, the story, or the question. You can directly copy and paste a span from the story or the question while answering the given question.", | |
| "task496-semeval-answer-generation": "Given news headlines, an edited word and funniness labels. The original sentence has word with the given format {word}. Create new headlines by replacing {word} in the original sentence with edit word. Your task is to generate answer \"Yes\" if an edited sentence matches funniness label, otherwise generate answer \"No\".", | |
| "task1214-atomic-classification-xwant": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether, as a result of the Head, PersonX wants what is mentioned in the Tail or not. In this task, wanting is a postcondition desire on the part of PersonX, respectively. As a result of PersonX giving PersonY gifts, PersonX may also desire to hug PersonY. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task1286-openbookqa-question-answering": "In this task, you are given a multiple-choice question and you have to pick the correct option. Answer with option indexes (i.e., \"A\", \"B\", \"C\", and \"D\").", | |
| "task219-rocstories-title-answer-generation": "In this task, you're given five sentences, numbered 1 through 5. Your job is to generate a title for the story that makes complete sense. The title must be short, with less than three words, use simple language, and include the main topic of the story.", | |
| "task155-count-nouns-verbs": "In this task, you need to count the number of nouns/verbs in the given sentence.", | |
| "task105-story-cloze-rocstories-sentence-generation": "In this task, you're given a four sentences of story written in natural language. Your job is to complete end part of the story by predicting appropriate last sentence which is coherent with the given sentences.", | |
| "task079-conala-concat-strings": "In this task you will be given a list of strings and you need to concatenate them.", | |
| "task1338-peixian-equity-evaluation-corpus-sentiment-classifier": "The input is a sentence. The sentence includes an emotion. The goal of the task is to classify the emotion in the sentence to one of the classes: 'fear', 'joy', 'anger', 'sadness'. The emotion mainly depends on the adverb within the sentence.", | |
| "task169-strategyqa-sentence-generation": "In this task, you are presented with a term, a question, and a yes/no answer. For context, the term is associated with a short description. Write the simple facts one needs to know in order to answer the given question. The facts should be easily found on the web (e.g. mentioned in a Wikipedia page)", | |
| "task1596-event2mind-text-generation-2": "The task is to generate text based off of an event. For the given event, there will be a person who does the action in the event (usually PersonX). Your task is to write that persons intent behind doing the event, i.e. why did the person do that action.", | |
| "task1540-parsed-pdfs-summarization": "In this task, you are given a part of an article. Your task is to generate headline (title) for this text. Preferred headlines are under fifteen words.", | |
| "task286-olid-offense-judgment": "In this task you are given a tweet and you must identify whether the tweet contains any offense or any form of (untargeted) profanity. Label the post as NOT if the post does not contain offense or profanity. Non-offensive posts do not include any form of offense or profanity. label the post as OFFENSIVE if the post contains offensive language or a targeted (veiled or direct) offense. Twitter user mentions were substituted by @USER and URLs have been substitute by URL.", | |
| "task220-rocstories-title-classification": "In this task, you're given five sentences, numbered 1 through 5, and two options a and b for possible titles for the story. Your job is to choose the title that better fits the story. Indicate your choice by 'a' or 'b'.", | |
| "task613-politifact-text-generation": "In this task, you are given a statement spoken by a politician in natural language. Your task is to generate the subject of the discussion for the given statement. The subject generated is not necessarily a part of the given input. Your answer should contain one or more words.", | |
| "task498-scruples-anecdotes-whoiswrong-classification": "Given a real-life anecdote of a complex ethical situation, identify who is wrong here. Classify each post into the following categories : AUTHOR (if the author of the post is wrong in the situation), OTHER (if any of the others mentioned in the paragraph is wrong), EVERYBODY (if everyone mentioned is wrong) and NOBODY (if no one is wrong in the situation).", | |
| "task420-persent-document-sentiment-classification": "Given a document and an entity the task is to select the author's sentiment towards the entity. Sentiments can be Positive, Neutral and Negative. Select Positive if the article expresses a positive view towards the given entity or praises its quality or skills. Select Neutral if the document expresses no clear view towards the entity or has equal amounts of positive and negative statements or expressing some fact/quote by someone else. Select Negative if the article expresses a negative view towards like harsh remarks, criticizing entities action/decision etc. Note that URLs in the text have been replaced with [Link].", | |
| "task233-iirc-link-exists-classification": "In this task, you're given a question, along with a context passage. The passage will not have a direct answer to the question, but can include some information and a clue for the answer. Assuming you can only obtain information about one entity in the passage, your job is to determine whether information from the passage can be used to answer the question. Indicate your choice as `a` for Yes or `b` for No.", | |
| "task1336-peixian-equity-evaluation-corpus-gender-classifier": "You will be given a sentence containing a pronoun/person name and an emotion. From these implicit parameters, the main goal is to find the gender of the person (male / female).", | |
| "task384-socialiqa-question-classification": "In this task, you're given a context passage, an answer, and a question. Your task is to classify whether the question for this answer is correct or not, based on the given context with commonsense reasoning about social situations. If its correct ,return \"True\" else \"False\".", | |
| "task002-quoref-answer-generation": "In this task, you're expected to write answers to questions involving multiple references to the same entity. The answer to the question should be unambiguous and a phrase in the paragraph. Most questions can have only one correct answer.", | |
| "task112-asset-simple-sentence-identification": "You will be given two pieces of text with the same meaning. One of them is simpler and easier to understand for non-native English speakers. Complex texts may contain more difficult words, have unnecessary phrases or contain long sentences. Your task is to choose the simpler piece of text. You are expected to output 'Text one' if the first sentence is simpler. Otherwise output 'Text two'.", | |
| "task584-udeps-eng-fine-pos-tagging": "In this task, you need to provide the parts-of-speech tag of a word present in a sentence specified within curly braces ( '{{ ... }}' ). The parts-of-speech tags are fine labels that represent a category of words with similar grammatical properties. The list of part-of-speech tags i.e tagset of this corpus is : '$': Dollar Sign, \"''\": Single Quotes, ',': Comma Symbol, '-LRB-': Left Parantheses, '-RRB-': Right Parantheses, '.': Period, ':': Colon, 'ADD': Email Address, 'AFX': Affix, 'CC': Coordinating conjunction, 'CD': Cardinal Number, 'DT': Determiner, 'EX': Existential there, 'FW': Foreign Word, 'GW': Go with, 'HYPH': Hyphen symbol, 'IN': Preposition or a subordinating conjunction, 'JJ': Adjective, 'JJR': A comparative Adjective, 'JJS': A Superlative Adjective, 'LS': List item Marker, 'MD': Modal, 'NFP': Superfluous punctuation, 'NN': Singular Noun, 'NNP': Singular Proper Noun, 'NNPS': Prural Proper Noun, 'NNS': Prural Noun, 'PDT': Pre-determiner, 'POS': Possessive Ending, 'PRP': Personal pronoun, 'PRP$': Possessive Pronoun, 'RB': Adverb, 'RBR': Comparative Adverb, 'RBS': Superlative Adverb, 'RP': Particle, 'SYM': Symbol, 'TO': To , 'UH': Interjection, 'VB': Base form Verb, 'VBD': Verb in Past tense, 'VBG': Verb in present participle, 'VBN': Verb in past participle, 'VBP': Verb in non-3rd person singular present, 'VBZ': Verb in 3rd person singular present, 'WDT': Wh-determiner, 'WP': Wh-pronoun, 'WP$' Possessive Wh-pronoun, 'WRB': Wh-adverb, 'XX': Unknown, '``': Double backticks.", | |
| "task1167-penn-treebank-coarse-pos-tagging": "Provide the parts-of-speech tag of a word present in a sentence specified within curly braces ( '{{ ... }}' ). The parts-of-speech tags are coarse labels that represent a category of words with similar grammatical properties. The list of part-of-speech tags i.e tagset of this corpus is - \n '.': Period symbol is used for symbols denoting Punctuations/Separations such as comma, period, backticks etc., \n 'ADJ': Adjectives are words that typically modify nouns and specify their properties or attributes, \n 'ADP': Adposition is a cover term for prepositions and postpositions, \n 'ADV': Adverbs are words that typically modify verbs for such categories as time, place, direction or manner, \n 'CONJ': A word used to connect clauses or sentences or to coordinate words in the same clause, \n 'DET': Determiners are words that modify nouns or noun phrases and express the reference of the noun phrase in context, \n 'NOUN': Nouns are a part of speech typically denoting a person, place, thing, animal or idea, \n 'NUM': A numeral is a word, functioning most typically as a determiner, adjective or pronoun, that expresses a number and a relation to the number, such as quantity, sequence, frequency or fraction, \n 'PRT': Particles are function words that must be associated with another word or phrase to impart meaning and that do not satisfy definitions of other universal parts of speech, \n 'PRON': Pronouns are words that substitute for nouns or noun phrases, whose meaning is recoverable from the linguistic or extralinguistic context, \n 'PROPN': A proper noun is a noun (or nominal content word) that is the name (or part of the name) of a specific individual, place, or object, \n 'VERB': A verb is a member of the syntactic class of words that typically signal events and actions, can constitute a minimal predicate in a clause, and govern the number and types of other constituents which may occur in the clause, \n 'X': The tag X is used for words that for some reason cannot be assigned a real part-of-speech category.", | |
| "task306-jeopardy-answer-generation-double": "You will be given a trivia clue, and the category it belongs to. You should answer with the best answer that belongs in the category and is described by the clue. For simplicity, answers should be in all lower cased letters.", | |
| "task069-abductivenli-classification": "In this task, you will be shown a short story with a beginning, two potential middles, and an ending. Your job is to choose the middle statement that makes the story coherent / plausible by writing \"1\" or \"2\" in the output. If both sentences are plausible, pick the one that makes most sense.", | |
| "task1168-brown-coarse-pos-tagging": "In this task, you need to provide the parts-of-speech tag of a word present in a sentence specified within curly braces ( '{{ ... }}' ). The parts-of-speech tags are coarse labels that represent a category of words with similar grammatical properties. The list of part-of-speech tags i.e tagset of this corpus is - \n '.': Period symbol is used for symbols denoting Punctuations/Separations such as comma, period, backticks etc., \n 'ADJ': Adjectives are words that typically modify nouns and specify their properties or attributes, \n 'ADP': Adposition is a cover term for prepositions and postpositions, \n 'ADV': Adverbs are words that typically modify verbs for such categories as time, place, direction or manner, \n 'CONJ': A word used to connect clauses or sentences or to coordinate words in the same clause, \n 'DET': Determiners are words that modify nouns or noun phrases and express the reference of the noun phrase in context, \n 'NOUN': Nouns are a part of speech typically denoting a person, place, thing, animal or idea, \n 'NUM': A numeral is a word, functioning most typically as a determiner, adjective or pronoun, that expresses a number and a relation to the number, such as quantity, sequence, frequency or fraction, \n 'PRT': Particles are function words that must be associated with another word or phrase to impart meaning and that do not satisfy definitions of other universal parts of speech, \n 'PRON': Pronouns are words that substitute for nouns or noun phrases, whose meaning is recoverable from the linguistic or extralinguistic context, \n 'PROPN': A proper noun is a noun (or nominal content word) that is the name (or part of the name) of a specific individual, place, or object, \n 'VERB': A verb is a member of the syntactic class of words that typically signal events and actions, can constitute a minimal predicate in a clause, and govern the number and types of other constituents which may occur in the clause, \n 'X': The tag X is used for words that for some reason cannot be assigned a real part-of-speech category.", | |
| "task103-facts2story-long-text-generation": "In this task, five ordered key facts are given. All the given facts are expressed in natural language. Your job is to generate a story 100 to 1000 words long, that includes all the facts given as input in their order of appearance while expanding upon them to produce a broader, yet coherent, narrative.", | |
| "task676-ollie-relationship-answer-generation": "Given a sentence and two mentions from the text (arguments), indicate a phrase (a verb or noun phrase) that describes the relationship between the provided arguments.", | |
| "task322-jigsaw-classification-threat": "In this task, you are given a public comment from online platforms. You are expected to classify the comment into two classes: threat and non-threat. Threat is a statement that someone will be hurt or harmed, especially if the person does not do something in particular.", | |
| "task428-senteval-inversion": "In this task you are given a sentence. You must judge whether there exist two consecutive words within the sentence with flipped orders, that is, whether the sentence will make sense and be correct if the order of two consecutive words changes. Label the instances as \"Inversion\" or \"Original\" based on your judgment.", | |
| "task295-semeval-2020-task4-commonsense-reasoning": "In this task, you are given an impractical statement. You are also given three reasons (associated with \"A\", \"B\", \"C\") explaining why this statement doesn't make sense. You must choose the most corresponding reason explaining why this statement doesn't make sense.", | |
| "task904-hate-speech-offensive-classification": "Classify the given tweet into the three categories: (1) 'Hate Speech', (2) 'Offensive' and (3) 'Neither'. 'Hate Speech' is kind of a threating statement or sometimes include call for violence while 'offensive' statement just offensds someone. 'Neither' is when it doesn't fall into Hate Speech or Offensive category.", | |
| "task196-sentiment140-answer-generation": "In this task, you are given a text from tweets and a boolean question whether this tweet has positive sentiment or negative sentiment. Your task is to generate answer \"yes\" when the tweet has that particular sentiment, otherwise generate answer \"no\".", | |
| "task210-logic2text-structured-text-generation": "In this task, you are given a natural language interpretation of commands (consist of logical operations) to select relevant rows from the given table. Your job is to generate command (in terms of logical operations) from given natural language interpretation. Define body (contains a collection of statements that define what the this logical operator does) of each logical operator between '{}' parenthesis. Here are the definitions of logical operators that you can use while generating command: \n 1. count: returns the number of rows in the view. \n 2. only: returns whether there is exactly one row in the view. \n 3. hop: returns the value under the header column of the row. \n 4. and: returns the boolean operation result of two arguments. \n 5. max/min/avg/sum: returns the max/min/average/sum of the values under the header column. \n 6. nth_max/nth_min: returns the n-th max/n-th min of the values under the header column. \n 7. argmax/argmin: returns the row with the max/min value in header column. \n 8. nth_argmax/nth_argmin: returns the row with the n-th max/min value in header column. \n 9. eq/not_eq: returns if the two arguments are equal. \n 10. round_eq: returns if the two arguments are roughly equal under certain tolerance. \n 11. greater/less: returns if the first argument is greater/less than the second argument. \n 12. diff: returns the difference between two arguments. \n 13. filter_eq/ filter_not_eq: returns the subview whose values under the header column is equal/not equal to the third argument. \n 14. filter_greater/filter_less: returns the subview whose values under the header column is greater/less than the third argument. \n 15. filter_greater_eq /filter_less_eq: returns the subview whose values under the header column is greater/less or equal than the third argument. \n 16. filter_all: returns the view itself for the case of describing the whole table \n 17. all_eq/not_eq: returns whether all the values under the header column are equal/not equal to the third argument. \n 18. all_greater/less: returns whether all the values under the header column are greater/less than the third argument. \n 19. all_greater_eq/less_eq: returns whether all the values under the header column are greater/less or equal to the third argument. \n 20. most_eq/not_eq: returns whether most of the values under the header column are equal/not equal to the third argument. \n 21. most_greater/less: returns whether most of the values under the header column are greater/less than the third argument. \n 22. most_greater_eq/less_eq: returns whether most of the values under the header column are greater/less or equal to the third argument.", | |
| "task195-sentiment140-classification": "In this task, you are given a text from tweets. Your task is to classify given tweet text into two categories: 1) positive, and 2) negative based on its content.", | |
| "task327-jigsaw-classification-toxic": "In this task, you are given a public comment from online platforms. You are expected to classify the comment into two classes: toxic and non-toxic. Toxicity is defiend as anything that is rude, disrespectful, or unreasonable that would make someone want to leave a converation.", | |
| "task308-jeopardy-answer-generation-all": "You will be given a trivia clue, and the category it belongs to. You should answer with the best answer that belongs in the category and is described by the clue. For simplicity, answers should be in all lower cased letters.", | |
| "task1504-hatexplain-answer-generation": "The input is a tweet which can be Hate Speech or Offensive. Given such a tweet, output a phrase from the tweet that makes it hate speech or offensive. The output should contain only one such phrase. The output has to be from within the tweet itself. Do not generate words or phrases not present in the tweet.", | |
| "task679-hope-edi-english-text-classification": "Classify the given comment into 'Hope Speech' or 'Not Hope Speech'. A hope speech is defined as content that is encouraging, positive or supportive contents and talks about equality, diversity or inclusion", | |
| "task583-udeps-eng-coarse-pos-tagging": "In this task, you need to provide the parts-of-speech tag of a word present in a sentence specified within curly braces ( '{{ ... }}' ). The parts-of-speech tags are coarse labels that represent a category of words with similar grammatical properties. The list of part-of-speech tags i.e. tagset of this corpus is 'ADJ': Adjectives are words that typically modify nouns and specify their properties or attributes, 'ADP': Adposition is a cover term for prepositions and postpositions, 'ADV': Adverbs are words that typically modify verbs for such categories as time, place, direction or manner, 'AUX': An auxiliary is a function word that accompanies the lexical verb of a verb phrase and expresses grammatical distinctions not carried by the lexical verb, such as person, number, tense, mood, aspect, voice or evidentiality, 'CCONJ': A coordinating conjunction is a word that links words or larger constituents without syntactically subordinating one to the other and expresses a semantic relationship between them, 'DET': Determiners are words that modify nouns or noun phrases and express the reference of the noun phrase in context, 'INTJ': An interjection is a word that is used most often as an exclamation or part of an exclamation, 'NOUN': Nouns are a part of speech typically denoting a person, place, thing, animal or idea, 'NUM': A numeral is a word, functioning most typically as a determiner, adjective or pronoun, that expresses a number and a relation to the number, such as quantity, sequence, frequency or fraction, 'PART': Particles are function words that must be associated with another word or phrase to impart meaning and that do not satisfy definitions of other universal parts of speech, 'PRON': Pronouns are words that substitute for nouns or noun phrases, whose meaning is recoverable from the linguistic or extralinguistic context, 'PROPN': A proper noun is a noun (or nominal content word) that is the name (or part of the name) of a specific individual, place, or object, 'PUNCT': Punctuation marks are non-alphabetical characters and character groups used in many languages to delimit linguistic units in printed text, 'SCONJ': A subordinating conjunction is a conjunction that links constructions by making one of them a constituent of the other. The subordinating conjunction typically marks the incorporated constituent which has the status of a (subordinate) clause, 'SYM': A symbol is a word-like entity that differs from ordinary words by form, function or both, 'VERB': A verb is a member of the syntactic class of words that typically signal events and actions, can constitute a minimal predicate in a clause, and govern the number and types of other constituents which may occur in the clause, 'X': The tag X is used for words that for some reason cannot be assigned a real part-of-speech category.", | |
| "task429-senteval-tense": "In this task you are given a sentence. You must judge whether the main verb of the sentence is in present or past tense. Label the instances as \"Present\" or \"Past\" based on your judgment. If there is no verb in the given text, answer \"Present\".", | |
| "task159-check-frequency-of-words-in-sentence-pair": "In this task, answer 'Yes' if the frequency of the given word in the two sentences is equal, else answer 'No'.", | |
| "task610-conllpp-ner": "In this task, you are given a text from a post. Your task is to find all of the proper nouns and label them. The labels are <B-PER>, <I-PER> for persons; <B-ORG>, <I-ORG> for organizations; <B-LOC>, <I-LOC> for locations; and <B-MISC>, <I-MISC> for other nouns. The capital 'B' denotes the first word of a proper noun phrase. The capital 'I' denotes all following words of the same noun phrase. If a word is not a part of a proper noun phrase, do not label it.", | |
| "task307-jeopardy-answer-generation-final": "You will be given a trivia clue, and the category it belongs to. You should answer with the best answer that belongs in the category and is described by the clue. For simplicity, answers should be in all lower cased letters.", | |
| "task500-scruples-anecdotes-title-generation": "Given a real-life anecdote of a complex ethical situation, generate a suitable title that describes the main event/root cause of the situation. Imagine that the given text is a social media post, and you have to write the title of the post, so the users can decide to read the main text or not.", | |
| "task115-help-advice-classification": "Read the given sentence and if it is a general advice then indicate via \"yes\". Otherwise indicate via \"no\". advice is basically offering suggestions about the best course of action to someone. advice can come in a variety of forms, for example Direct advice and Indirect advice. (1) Direct advice: Using words (e.g., suggest, advice, recommend), verbs (e.g., can, could, should, may), or using questions (e.g., why don't you's, how about, have you thought about). (2) Indirect advice: contains hints from personal experiences with the intention for someone to do the same thing or statements that imply an action should (or should not) be taken.", | |
| "task1384-deal-or-no-dialog-classification": "Given a negotiation between two participants, answer 'Yes' if both participants agree to the deal, otherwise answer 'No'.", | |
| "task178-quartz-question-answering": "You are given a question or fill-in-the-blank question, two answer options (Option1 and Option2) and an Explanation. Your task is to find the correct answer (return the string of the correct option, not option1/2) for the given question from the given options and using explanation.", | |
| "task067-abductivenli-answer-generation": "In this task, you're given the beginning and the ending of a three-part story. Your job is to complete the short story by writing a middle sentence that seamlessly connects the first and last sentence. Generated sentences must be short, have fewer than 10 words, and be simple as if narrating to a child. Avoid using any irrelevant extra information when creating the middle sentence, and use names (e.g. Jack, Barbara,...) instead of pronouns (e.g. he / she) wherever possible.", | |
| "task1196-atomic-classification-oeffect": "In this task, you are given a tuple, comprising Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether, as a result of the Head, PersonY or others will be affected as mentioned in the Tail or not. These effects are social actions that may occur after the Head event. For example, X may get hugged, and Y may blush in response. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task106-scruples-ethical-judgment": "In this task, you are given two simple actions (associated with \"A\", \"B\"). You must identify which action is considered less ethical. Do not generate anything else apart from one of the following characters: 'A', 'B'.", | |
| "task1661-super-glue-classification": "In this task, you are given Wikipedia articles on a range of topics as passages and a question from the passage. We ask you to answer the question by classifying the answer as 0 (False) or 1 (True)", | |
| "task1285-kpa-keypoint-matching": "The input contains a debate topic, an argument on the topic and a keypoint, separated by \"<sep>\". Your task is to answer if the keypoint matches the argument and summarizes exactly what the argument means, in the context of the given topic.", | |
| "task235-iirc-question-from-subtext-answer-generation": "In this task, you're given a context, further information available on a particular linked term from the statement, and an answer term. Your job is to generate a question that can use the information provided to obtain the given answer. You should use the information on both context and link information to create the question. Note that the answer to the question should be exactly the given answer, and if the answer is none, the answer to the question shouldn't be obtainable from the context or linked information.", | |
| "task132-dais-text-modification": "Languages typically provide more than one grammatical construction to express certain types of messages. Your task is to generate a senetence with the same meaning as given sentence, but with different grammatical construction.", | |
| "task049-multirc-questions-needed-to-answer": "You are given a paragraph, with numbered sentences, and a question. Write the numbers of the sentences needed to answer the given question. Separate multiple sentence numbers with a comma (\", \"). E.g. \"1, 2\" would be the correct response if Sent 1 and Sent 2 are needed to answer the given question (do not include sentences that contain no information that is necessary to answer the question). Sort the sentence numbers in an increasing order (i.e., '1, 2', instead of '2, 1'). Your answer should not contain anything than the sentence numbers separated by a comma (e.g. '1, 2, 3, 4'). The given question requires more than one sentence to answer. So, your answer must contain at least one comma (\", \").", | |
| "task157-count-vowels-and-consonants": "In this task, you need to count the number of vowels (letters 'a', 'e', 'i', 'o', 'u') / consonants (all letters other than vowels) in the given sentence.", | |
| "task1587-scifact-classification": "In this task, you are given a abstract of article and corresponding title of an article. Your task is to generate label \"yes\" if title is right for article, otherwise generate \"no\".", | |
| "task040-qasc-question-generation": "Turn the given fact into a question by a simple rearrangement of words. This typically involves replacing some part of the given fact with a WH word. For example, replacing the subject of the provided fact with the word \"what\" can form a valid question. Don't be creative! You just need to rearrange the words to turn the fact into a question - easy! Don't just randomly remove a word from the given fact to form a question. Remember that your question must evaluate scientific understanding. Pick a word or a phrase in the given fact to be the correct answer, then make the rest of the question. You can also form a question without any WH words. For example, \"A radio converts electricity into?\"", | |
| "task316-crows-pairs-classification-stereotype": "In this task, you are given a short passage that conveys a stereotype or an anti-stereotype. A stereotype is an over-generalized belief about a particular group of people(based on race, ethnicity, gender, age, abilities, etc.). An anti-stereotype is an idea that goes against a common stereotype. You are expected to classify the passage into two classes: stereotype or anti-stereotype.", | |
| "task1542-every-ith-element-from-starting": "In this task, you are given an input i,A where i is an integer and A is an array. You need to find every ith element of A starting with the 1st element.", | |
| "task929-products-reviews-classification": "Given an English language product review, determine if it is a Good Review or a Bad Review. A good review is one where the product's review has positive tone and Bad review is one where the tone of product's review is negative.", | |
| "task160-replace-letter-in-a-sentence": "In this task, you need to replace a letter in the sentence with another given letter.", | |
| "task747-glucose-cause-emotion-detection": "In this task, you will be given a short story. One sentence from the story is chosen. Consider the likely emotions and basic human drives of the participants in that sentence. Does any of these states of mind/feelings motivate the participant to do what happens in that sentence? You should write your answer in the form \" A >Motivates> B\". Try to use phrases and sentences from the story to compose your answer when possible. For the motivation sentence, you must choose a verb from :feel(s), want(s) or like(s). There will always be some motivation in the given story.", | |
| "task202-mnli-contradiction-classification": "In this task, you're given a statement, and three sentences as choices. Your job is to determine which sentence clearly disagrees with the statement. Indicate your answer as '1', '2', or '3' corresponding to the choice number of the selected sentence.", | |
| "task1195-disflqa-disfluent-to-fluent-conversion": "Given a disfluent sentence, modify the sentence to it to its equivalent fluent form, preserving the meaning of the sentence.", | |
| "task305-jeopardy-answer-generation-normal": "You will be given a trivia clue, and the category it belongs to. You should answer with the best answer that belongs in the category and is described by the clue. For consistency, answers should be in all lower cased letters.", | |
| "task216-rocstories-correct-answer-generation": "In this task, you're given the title of a five-sentence story and the first four sentences. Your job is to write the last sentence of the story such that it seamlessly connects with the rest of the story.", | |
| "task574-air-dialogue-sentence-generation": "In this task, you're given a dialogue between a customer and a flight booking agent with a gap in the conversation. Your job is to find the answer of the previous dialogue. Avoid using irrelevant extra information while creating the answer. The answer should be relevant to the question before the blank. If you fill the blank with a question, it should have an answer from the agent in the given dialogue. Fill the gap marked with underline.", | |
| "task1360-numer-sense-multiple-choice-qa-generation": "Given a sentence with a missing word, pick the answer option that best fills out the missing word in the sentence. Indicate each answer with its index ('a', 'b', 'c', 'd').", | |
| "task037-qasc-generate-related-fact": "Write a fact related to the given fact, based on the given topic word. Note that, your fact should have at least one word in common with the given fact. All facts in this task refer to scientific facts. Your related fact must form a chain with the given fact. Chains form when two facts connect together to produce the third fact. An example of a chain is: \"pesticides cause pollution\" (given fact) + \"pollution can harm animals\" (related fact) → \"pesticides can harm animals\" (connected chain). Avoid creating simple paraphrases of the given fact. While your generated fact should be related to the input fact, they must describe slightly different scientific phenomena. It's okay if your related fact includes some irrelevant information, provided that it has some overlap with the given fact and it contains some words corresponding to the provided topic.", | |
| "task035-winogrande-question-modification-person": "You're given a fill-in-the-blank question where the answer is PersonX. You need to minimally change the given question so that the answer flips to PersonY. This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from \"sympathetic\" to \"stern\"). You should not change any content in the given question beyond a word or two i.e. the trigger word/phrase. PersonX and PersonY should not be equally likely to fill the blank. For your question, PersonY should be a well-agreed answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. Do not use the names of real people or generic names (e.g., Donald Trump, John Doe, etc.) in your question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. Make sure that PersonX and PersonY have the same gender. In your question, PersonX and PersonY should be used only ONCE and PersonX should appear earlier than PersonY. Although there are many correct answers, you only need to write one of them.", | |
| "task898-freebase-qa-answer-generation": "Given an trivia question precisely answer the question with a word/phrase/name. External resources such as Wikipedia could be used to obtain the facts.", | |
| "task380-boolq-yes-no-question": "In this task you will be given a passage and a yes/no question based on the passage. You should answer the question using the information from the passage.", | |
| "task1678-mathqa-answer-selection": "Given a math problem with context and a question and 5 answer choices, the task is to provide the correct answer choice based on the problem. You must choose one of the given answer choices by letter: a, b, c, d, or e; anything else is invalid.", | |
| "task1187-politifact-classification": "In this task, you are given a statement spoken by a politician in natural language and the subject of the discussion for the given statement. Your task is to determine if the given subject of the discussion is correct for the given statement. The subject can be multiple words. Answer with \"True\" if it's the correct subject of discussion or return \"False\" if it's incorrect.", | |
| "task360-spolin-yesand-response-generation": "Given the sentence, generate \"yes, and\" response. \"Yes, and\" is a rule-of-thumb in improvisational comedy that suggests that a participant in a dialogue should accept what another participant has stated (\"Yes\") and then expand on that line of thought or context (\"and...\"). 1 In short, a \"Yes, and\" is a dialogue exchange in which a speaker responds by adding new information on top of the information/setting that was constructed by another speaker. Note that a \"Yes, and\" does not require someone explicitly saying 'yes, and...' as part of a dialogue exchange, although it could be the case if it agrees with the description above. There are many ways in which a response could implicitly/explicitly agree to the prompt without specifically saying 'yes, and...'.", | |
| "task1204-atomic-classification-hinderedby": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether the Head can be hindered by what is mentioned in the Tail or not. In this task, hindering introduces hindrances that obstruct the natural path to the achievement of a goal. For example, the event PersonX adopts a cat can be obstructed if PersonX is allergic to cats. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task517-emo-classify-emotion-of-dialogue": "In this task you will be given some text dialogue and you need to infer the underlying emotion of the text. The possible emotions are happy, sad, angry, or other.", | |
| "task494-review-polarity-answer-generation": "Given a review text from amazon and its polarity (either positive or negative). Generate answer \"True\" if given sentence and its polarity match, otherwise generate answer \"False\". A review is considered positive if the reviewer is satisfied with the product. Otherwise, it is considered negative.", | |
| "task1712-poki-classification": "You are given a short poem which is written by a kid. Your task is to predict if the kid is in elementary school or high school. Anyone who is from grade 1st to 6th-grade will be considered as in elementary school, and 7th to 12th-grade kids will be considered as in high school. There are only two possible outputs, i.e., elementary and high. All inputs have at least one output and it cannot have both categories at a time.", | |
| "task847-pubmedqa-question-generation": "Given a passage, construct a question on the basis of the information present in the passage. Construct the question in such a way that (i) it is unambiguous, (ii) it is answerable from the passage, (iii) its answer is unique (iv) it answer uses text phrases from the passage. Avoid creating questions that (i) can't be answered correctly without actually understanding the passage and (ii) uses the same words or phrases given in the passage.", | |
| "task1392-superglue-multirc-answer-verification": "In this task, you will be presented with a context passage, a question about that paragraph, and a possible answer to that question. The task is to check the validity of the answer. Answer with \"Yes\" or \"No\".", | |
| "task077-splash-explanation-to-sql": "In this task you are expected to provide an SQL statement from an english description of what that SQL statement does. The description may include multiple steps but you should only ouput one SQL statement that accomplishes every step. An SQL query works by selecting data from a table where certain conditions apply. A table contains columns where every row in that table must have a value for each column. Every table has a primary key that uniquely identifies each row, usually an id. To choose which columns are returned you specify that after the \"SELECT\" statement. Next, you use a \"FROM\" statement to specify what tables you want to select the data from. When you specify a table you can rename it with the \"AS\" statement. You can reference that table by whatever name follows the \"AS\" statement. If you want to select data from multiple tables you need to use the \"JOIN\" statement. This will join the tables together by pairing a row in one table with every row in the other table (Cartesian Product). To limit the number of rows returned you should use the \"ON\" statement. This will only return rows where the condition specified after the statement is true, this is usually an equals operator with primary keys. You can also use the \"WHERE\" statement to specify that only rows with column values statisfying a certain condition, should be returned. The \"GROUP BY\" statement will group rows together that have equal column values for whatever columns follows the statement. The \"HAVING\" statement will return groups that statisfy whatever condition follows the statement. Any column(s) being returned from grouped rows must either be an aggregate function, (AVG, MAX, COUNT, SUM, ...) of a column, or the column(s) that the data was grouped by. To sort the returned data you can use the \"ORDER BY\" command which will order the data by whatever aggregate function or column follows the statement. The \"DESC\" statement will sort in descending order and the \"ASC\" statement will sort in ascending order. Finally, you can use the \"LIMIT\" statement to return a certain number of rows. When \"*\" is used in an SQL statement every column is returned. For example, SELECT * FROM table WHERE attribute = 1, will select every column from rows with the attribute column equal to 1.", | |
| "task182-duorc-question-generation": "In this task, you will be shown an extract from a movie plot. You need to read the extract and create questions that can be answered from the extract. Once you create a question you also need to provide the correct answer for it. Write your answer like \"Question: your question, Answer: your answer\". The questions should be well-formed and grammatically correct. The questions should be completely answerable from the given passage and should not require any external knowledge. Subjective questions are not allowed. Short answers containing words that are present in the passage are preferred. Create questions that result in factoid answers. A simple rule of thumb to decide whether a question is a factoid or not is to see if two different people with average reading/comprehension skills would come up with the same answer after reading the passage.", | |
| "task578-curiosity-dialogs-answer-generation": "In this task, you are given dialogue, and you have to find the most critical location in the given conversation.", | |
| "task234-iirc-passage-line-answer-generation": "In this task, you're given a question, along with a context passage which has extra information available on certain terms mentioned in it. Your job is to determine which sentence(s) from the passage can be used to search for further information needed to answer the question.", | |
| "task568-circa-question-generation": "In this task, you are given an answer, and your task is to generate a reasonable question for that answer.", | |
| "task239-tweetqa-answer-generation": "In this task, you are given a context paragraph of the tweet and question. Your task is to generate right answer of given question based on given context tweet paragraph.", | |
| "task177-para-nmt-paraphrasing": "This is a paraphrasing task. In this task, you're given a sentence and your task is to generate another sentence which express same meaning as the input using different words.", | |
| "task1202-atomic-classification-xneed": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether PersonX needs what is mentioned in the Tail before the Head or not. Here need is a precondition for PersonX to achieve the event. For example, in order for PersonX to give PersonY gifts, PersonX must first buy the presents. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task669-ambigqa-answer-generation": "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.", | |
| "task153-tomqa-find-location-hard-clean": "Given a story, answer the question about the story. The question is the last sentence in the input. These stories can be difficult due to their length and how each story has at least one of the three following scenarios: the first is when the individual's belief matches reality, the second is when the individual's belief does not match reality, and the third is when an individual has a false belief about another individual's beliefs. The question will ask about the location of an object in the story with respect to either none or one of the three scenarios.", | |
| "task217-rocstories-ordering-answer-generation": "In this task, you're given the title of a story consisting of five sentences, numbered 1 through 5. Your job is to arrange the sentences in order to make a story that makes complete sense and is apt for the title. Indicate your answer using the number of the sentences in order, such as '34152'.", | |
| "task1729-personachat-generate-next": "Your task is to generate the next utterance in a given dialogue. You will be given a few sentences describing the personality of the person who is making the dialogue, and a history of the dialogue after that. Each line in the history is said by one of the two participants in the conversation.", | |
| "task649-race-blank-question-generation": "In this task, you're given an article and an answer. Your task is to generate the fill-in-the-blank question for this answer based on the given article. The blanks are always at the end of the question, and can be filled with more than one word or phrase.", | |
| "task580-socialiqa-answer-generation": "In this task, you're given a context, a question, and three options. Your task is to find the correct answer to the question using the given context and options. Also, you may need to use commonsense reasoning about social situations to answer the questions. Classify your answers into 'A', 'B', and 'C'.", | |
| "task292-storycommonsense-character-text-generation": "In this task, you're given a story (which contains five sentences only). Your task is to find all the characters which are available in the given story.", | |
| "task572-recipe-nlg-text-generation": "In this task, you will be presented with the directions of a recipe separated by \",\" and have to fill in the \"___\" which is a step that is missing from the recipe.", | |
| "task088-identify-typo-verification": "The given sentence contains a typo which could be one of the following four types: (1) swapped letters of a word e.g. 'niec' is a typo of the word 'nice'. (2) missing letter in a word e.g. 'nic' is a typo of the word 'nice'. (3) extra letter in a word e.g. 'nicce' is a typo of the word 'nice'. (4) replaced letter in a word e.g 'nicr' is a typo of the word 'nice'. You need to identify the typo in the given sentence. To do this, answer with the word containing the typo.", | |
| "task361-spolin-yesand-prompt-response-classification": "Given the prompt and a response, classify the them to \"yes\" if response is \"yes, and\" type. Otherwise classify it as \"no\". \"Yes, and\" is a rule-of-thumb in improvisational comedy that suggests that a participant in a dialogue should accept what another participant has stated (\"Yes\") and then expand on that line of thought or context (\"and...\"). 1 In short, a \"Yes, and\" is a dialogue exchange in which a speaker responds by adding new information on top of the information/setting that was constructed by another speaker. Note that a \"Yes, and\" does not require someone explicitly saying 'yes, and...' as part of a dialogue exchange, although it could be the case if it agrees with the description above. There are many ways in which a response could implicitly/explicitly agree to the prompt without specifically saying 'yes, and...'.", | |
| "task1726-mathqa-correct-answer-generation": "The task is to generate the correct answer for a given math problem.", | |
| "task1408-dart-similarity-classification": "This task is about classifying the similarity of two sentences. The sentences can be classified as (a) SIMILAR - similar to each other, and (b) DISSIMILAR - not similar to each other. Sentences that have the same RDF relationship in terms of [subject, predicate, object] are similar to each other. The input is a list of two sentences and the output is either SIMILAR or DISSIMILAR.", | |
| "task164-mcscript-question-answering-text": "You are given a paragraph (Passage), a question (Question) and two answer options (Option1 and Option2). Your task is to find the correct answer (and return the string of the correct option, not option1/2) for the given question from the given options and based on the given passage. Answer of the question can be found directly from the passage.", | |
| "task294-storycommonsense-motiv-text-generation": "In this task, you're given a context, a sentence, and a character. The sentence describes an action or job of the given character. Also, the context provides more information about the sentence or the character. Your task is to write the character's motivation by doing a specific job, which is given in the sentence. You can use context to find the motivation; however, all the inputs do not provide context. Also, in some inputs, there can be zero motivation; output, 'None' in this case.", | |
| "task1345-glue-qqp-question-paraprashing": "In this task you're given a question and you have to paraphrase the question to create the output question while retaining the meaning of the original question.", | |
| "task1205-atomic-classification-isafter": "In this task, you are given two phrases: Head and Tail, separated with <sep>. The Head and the Tail events are short phrases possibly involving participants. The names of specific people have been replaced by generic words (e.g., PersonX, PersonY, PersonZ). PersonX is always the subject of the event. You have to determine whether the Head happens after the Tail or not. Classify your answers into \"Yes\" and \"No\". The phrase may also contain \"___\", a placeholder that can be an object, a person, and/or an action.", | |
| "task515-senteval-odd-word-out": "In this task, you are given a sentence. You must judge whether a single noun or verb has been replaced with another word with the same part of speech. The inversion would result in the sentence sounding unnatural, So unnatural sentences will be considered changed. Label the instances as \"Original\" or \"Changed\" based on your judgment.", | |
| "task330-gap-answer-generation": "In this task, you will be presented with a text and a pronoun. You should write the name that the pronoun refers to. Position of the pronoun in the text is showed within two \"_\"s.", | |
| "task339-record-answer-generation": "In this task, you will be presented with a passage and have to answer a question based on your understanding of the events inferred from the passage. Try to find the best answer that is most likely to fill in \"_\". Note that the URLs in the text have been replaced with [Link].", | |
| "task1536-daily-dialog-happiness-classification": "You are given a conversation between two people. 'Person1:' and 'Person2:' are used to separate their respective dialogues. Your task is to classify the conversation either convey 'No emotion' or 'Happiness' by providing '1' and '0', respectively.", | |
| "task615-moviesqa-answer-generation": "You are given an open-domain question from an open movie database. Your task is to provide an answer to that question. Try to only include the answer. Do not put it in a sentence.", | |
| "task402-grailqa-paraphrase-generation": "Compose a new way to ask the same question. The new question must be asking the exact same thing as the original question does. Without any other information, you should be able to tell that the original question and the new question are the same. Preserve bracketed phrases in your new question. You can change the rest of the question provided the above requirements are satisfied. The new question must have a natural, fluent formulation, as if it were composed by a native English speaker. You will be given three pieces of information to help you create better questions. Those are: \"Answer\": the answer to the question, e.g., \"Honolulu\". \"Answer Type\": the type of the answer, e.g., \"Location\". \"Auxiliary Information\" is a detailed description of some concepts mentioned in the question, if any.", | |
| "task1356-xlsum-title-generation": "Generate an appropriate title for the given text. The generated title must be short and include the main topic of the text. The preferred titles are under fifteen words.", | |
| "task293-storycommonsense-emotion-text-generation": "In this task, you're given a context, a sentence, and a character. The sentence describes an action or job of the given character. Also, the context provides more information about the sentence or the character. Your task is to return one of the emotions which are expressed by the Character in the given sentence. For that you can use the Context; however, Context is not provided in all the inputs. Also, in some inputs, there can be zero emotion; for that return 'None'.", | |
| "task197-mnli-domain-answer-generation": "In this task, you're given a pair of sentences, sentence 1 and sentence 2. Your job is to write a single word that describes the genre that the two sentences belong to, such as face-to-face, government, letters, 9/11, slate, telephone, travel, verbatim, oup, fiction.", | |
| "task188-snli-neutral-to-entailment-text-modification": "In this task, you're given a pair of sentences, sentence 1 and sentence 2, that neither agree with nor contradict each other. Your job is to alter sentence 2 so that the pair clearly agree with each other. Generated sentences must be short, with less than 15 words. New information can be introduced. Avoid using pronouns to confuse the subject of the sentence.", | |
| "task453-swag-answer-generation": "Given a sentence, generate what should be the most likely next statement. The next statement should be reasonable and logically correct.", | |
| "task1292-yelp-review-full-text-categorization": "You are given a review about a place. You need to provide a rating from \"1 star\" to \"5 stars\" for this place.", | |
| "task571-recipe-nlg-ner-generation": "In this task, you have to identify the named entities (NER) which are the ingredients required given its directions. Named entities are the names of the items without their quantity.", | |
| "task518-emo-different-dialogue-emotions": "In this task you will be given two dialogues. You need to determine if both dialogues have the same underlying emotion. The possible emotions are happy, sad, angry, or other. If they do output 'yes', if not output 'no'.", | |
| "task329-gap-classification": "In this task, you will be presented with a text, a pronoun from the text, and two candidate names. You should determine what the pronoun refers to and classify the answers into A, B, or Neither. A and B here are referring to option A and option B. Position of the pronoun in the text is showed within two \"_\"s.", | |
| "task670-ambigqa-question-generation": "In this task, you're given an ambiguous question (which can be answered in more than one way). Your task is to write a question that clarifies the given question in such a way that the generated question has one unique answer.", | |
| "task595-mocha-answer-generation": "In this task, you're given a context passage, followed by a question that needs to be answered. Based on the paragraph, you must write unambiguous answers to the questions and your answer must refer to a specific phrase from the paragraph. If multiple answers seem to exist, write the answer that is the most plausible.", | |
| "task089-swap-words-verification": "In this task, positions of two consecutive words have been swapped. You need to output the position of the swapped words. e.g. in the sentence 'Seattle has nice a weather', the word 'a' and 'nice' have been swapped. These words are at 3rd and 4th position respectively. So, the output should be (3, 4). Note that the index of first word is 1.", | |
| "task591-sciq-answer-generation": "Given a scientific question, generate a correct answer to it.", | |
| "task237-iirc-answer-from-subtext-answer-generation": "In this task, you're given a statement, further information available on a particular linked term from the statement, and a question. Your job is to generate the answer to the question by using the information provided. If there is no clear answer obtainable, output 'none'.", | |
| "task238-iirc-answer-from-passage-answer-generation": "In this task, you're given a passage, further information available on a particular linked term from the statement, and a question. Your job is to generate the answer to the question by using the information provided. If there is no clear answer obtainable, output 'none'.", | |
| "task431-senteval-object-count": "In this task you are given a sentence. You must judge whether the object of the main clause is singular(like: apple) or plural(like: apartments). Label the instances as \"Singular\" or \"Plural\" based on your judgment.", | |
| "task1325-qa-zre-question-generation-on-subject-relation": "You will be given a context, a subject and a relation. Your task is to generate a question based on the subject and relation. The generated question should include the given subject. Try to use a minimum number of words that are not present in either context, subject or relation while generating question.", | |
| "task285-imdb-answer-generation": "In this task, you are given a review of a movie and a boolean question whether this review has positive sentiment or negative sentiment. Your task is to generate answer \"yes\" when the tweet has that particular sentiment, otherwise generate answer \"no\".", | |
| "task1290-xsum-summarization": "In this task, you are given an article. Your task is to summarize the article in a sentence.", | |
| "task900-freebase-qa-category-classification": "Given a trivia question, classify broad topical category from this list: 'theater', 'geology', 'book', 'tv', 'astronomy', 'aviation', 'military', 'government', 'boxing', 'projects', 'metropolitan_transit', 'law', 'venture_capital', 'broadcast', 'biology', 'people', 'influence', 'baseball', 'spaceflight', 'media_common', 'cvg', 'opera', 'olympics', 'chemistry', 'visual_art', 'conferences', 'sports', 'language', 'travel', 'location', 'award', 'dining', 'martial_arts', 'comic_strips', 'computer', 'user', 'tennis', 'music', 'organization', 'food', 'event', 'transportation', 'fictional_universe', 'measurement_unit', 'meteorology', 'distilled_spirits', 'symbols', 'architecture', 'freebase', 'internet', 'fashion', 'boats', 'cricket', 'film', 'medicine', 'finance', 'comic_books', 'celebrities', 'soccer', 'games', 'time', 'geography', 'interests', 'common', 'base', 'business', 'periodicals', 'royalty', 'education', 'type', 'religion', 'automotive', 'exhibitions'.", | |
| "task141-odd-man-out-classification-category": "Given a category and a set of five words, find the word from the set that does not belong (i.e. is the least relevant) with the other words in the category. Words are separated by commas.", | |
| "task236-iirc-question-from-passage-answer-generation": "In this task, you're given a passage, further information available on a particular linked term from the statement, and an answer term. Your job is to generate a question that can use the information provided to obtain the given answer. You should use the information on both passage and link information to create the question. Note that the answer to the question should be exactly the given answer, and if the answer is none, the answer to the question shouldn't be obtainable from the passage or linked information.", | |
| "task230-iirc-passage-classification": "In this task, you're given a question, along with three passages, 1, 2, and 3. Your job is to determine which passage can be used to answer the question by searching for further information using terms from the passage. Indicate your choice as 1, 2, or 3.", | |
| "task1340-msr-text-compression-compression": "Given a text, write a compressed version of it in a single sentence.", | |
| "task750-aqua-multiple-choice-answering": "You are given a math word problem and you are supposed to make use of mathematical concepts like probability, combinatorics, linear equations, sequences/series, or simple operators like addition, subtraction, division, or multiplication to answer the given question. Choose the best option from the 5 given options and report that option.", | |
| "task240-tweetqa-question-generation": "In this task, you are given a context tweet and an answer. Your job is to generate a question for the given answer based on the given tweet paragraph. Note that your question should be answerable based on the given tweet, and the answer to your question should be the given answer.", | |
| "task289-gigaword-summarization": "In this task, you are given a text of article and corresponding title of an article. Your task is to generate label \"yes\" if headline (title) is right for article, otherwise generate \"no\".", | |
| "task348-squad2.0-unanswerable-question-generation": "This task is about creating an unanswerable question based on a given passage. Construct a question that looks relevant to the given context but is unanswerable. Following are a few suggestions about how to create unanswerable questions:\n(i) create questions which require satisfying a constraint that is not mentioned in the passage\n(ii) create questions which require information beyond what is provided in the passage in order to answer\n(iii) replace an existing entity, number, date mentioned in the passage with other entity, number, date and use it in the question\n(iv) create a question which is answerable from the passage and then replace one or two words by their antonyms or insert/remove negation words to make it unanswerable.", | |
| "task200-mnli-entailment-classification": "In this task, you're given a statement and three sentences as choices. Your job is to determine which sentence can be inferred from the statement. Incorrect choices change the meaning in important ways or have details that are not mentioned in the statement. Indicate your answer as 1,2, or 3 corresponding to the choice number of the selected sentence.", | |
| "task1502-hatexplain-classification": "The input is a tweet which can be classified as Hate Speech, Offensive or Normal. Given such a tweet, output the class the tweet belongs to. Hate Speech will contain threatening language towards the community targeted. Offensive language will contain abusive or discriminatory language towards the community targeted.", | |
| "task038-qasc-combined-fact": "Combine the given two facts to write a concluding fact. Note that there should be some parts of the first and second facts that are not mentioned in this conclusion fact. Your combined fact should be the result of a chain between the two facts. Chains form when two facts connect together to produce a concluding fact. An example of a chain is: \"pesticides cause pollution\" (fact1) + \"pollution can harm animals\" (fact2) → \"pesticides can harm animals\" (conclusion fact). Parts of the concluding fact overlap with the first and the second fact. To construct such a concluding fact, a useful way is to borrow the subject from one fact and the ending conclusions from another fact.", | |
| "task834-mathdataset-classification": "You are supposed to identify the category of a high-school level math question. There are five possible categories (1) algebra (2) arithmetic (3) measurement (4) numbers, and (5) probability. Use the following guidelines: (1) 'algebra' questions will typically contain letter variables and will ask you to find the value of a variable (2) 'arithmetic' questions will ask the sum, difference, multiplication, division, power, square root or value of expressions involving brackets (3) 'measurement' questions are questions that ask to convert a quantity from some unit to some other unit (4) 'numbers' questions will be about bases, remainders, divisors, GCD, LCM etc. (5) 'probability' questions will ask about the probability of the occurrence of something. A question must belong to only one category.", | |
| "task1554-scitail-classification": "In this task, you are given two statements. The task is to output whether a given textual premise, i.e. Statement 2, entails or implies a given scientific fact, i.e. Statement 1. The output should be 'entails' if Statement 2 supports Statement 1 and should be 'neutral' otherwise.", | |
| "task820-protoqa-answer-generation": "Write a correct answer for the question. You should reason about prototypical situations and provide the correct answer. Avoid answers that provide incomplete justification for the question.", | |
| "task344-hybridqa-answer-generation": "In this task, you will be presented with a question and you have to answer the question based on your knowledge. Your answers should be as short as possible.", | |
| "task066-timetravel-binary-consistency-classification": "In this task, you are given a short story consisting of exactly 5 sentences where the second sentence is missing. You are given a candidate for the second sentence and you need to identify if the given sentence connects the first sentence with the rest of the story. Indicate your answer by \"Yes\" if it connects, otherwise \"No\". Do not generate anything else apart from \"Yes\" or \"No\". The given sentence is incorrect if it changes the subsequent storyline, so that at least one of the three subsequent sentences form a consistent story.", | |
| "task1347-glue-sts-b-similarity-classification": "Evaluate the similarity between them and classify them into classes from 0-5 as follows:\n0 : The two sentences are completely dissimilar.\n1 : The two sentences are not equivalent, but are on the same topic.\n2 : The two sentences are not equivalent, but share some details.\n3 : The two sentences are roughly equivalent, but some important information differs/missing.\n4 : The two sentences are mostly equivalent, but some unimportant details differ.\n5 : The two sentences are completely equivalent, as they mean the same thing.", | |
| "task823-peixian-rtgender-sentiment-analysis": "Given a 'poster' sentence and a corresponding 'response' (often, from Facebook or Reddit)classify the sentiment of the given response into four categories: 1) Positive, 2) Negative, 3) Neutral, and 4) Mixed if it contains both positive and negative.", | |
| "task875-emotion-classification": "In this task, you are given a sentence containing a particular emotion. You must classify the sentence into one of the six emotions: 'joy', 'love', 'anger', 'fear', or 'surprise'.", | |
| "task575-air-dialogue-classification": "In this task, you are given a dialogue from a conversation between an agent and a customer. Your task is to determine the speaker of the dialogue. Answer with \"agent\" or \"customer\".", | |
| "task074-squad1.1-question-generation": "This task is about reading the given passage and construct a question about the information present in the passage. Construct a question in such a way that (i) it is unambiguous, (ii) it is answerable from the passage, (iii) its answer is unique (iv) its answer is a continuous text span from the paragraph. Avoid creating questions that (i) can be answered correctly without actually understanding the paragraph and (ii) uses same words or phrases given in the passage.", | |
| "task247-dream-answer-generation": "In this task, you will be shown a conversation and a question. You need to answer the question and choose the correct option based on the conversation. \"W\" and \"M\" in the conversations stand for \"woman\" and \"man\".", | |
| "task176-break-decompose-questions": "In this task you will break down a question into the basic steps required to answer it.\n A question decomposition is a numbered list of operations that must be performed to answer the original question. Imagine explaining your question to a friendly droid by listing each action it should take in order for the question to be answered. Each step in our decomposition should refer to either an entity (known or unknown), a propery of an entity or a query operation (count, group, union, etc.)\n Here are the list of step templates and their description:\n Select: A select step is used to return a set of objects. There are no references to previous steps in a select step. template: Return [attributes]\n Filter: A filter step is used to return results from a previous step to which a certain condition applies. template: Return [#step] [condition]\n Project: A project step should return certain attributes of the results of a previous step. template: Return [attributes] of [#step]\n Aggregate: An aggregate step returns an aggregator function applied on a step's result. template: Return the [aggregator] of [#step].\n Group: A group step is an aggregator applied on attributes. template: Return the [aggregator] of [#step] for each [attribute]\n Superlative: A superlative step is used to return the result with a highest/lowest attribute among other results. template: Return [#step1] [where] [#step2] [is] [highest / lowest]\n Comparative: A comparative step is used when we need to compare an attribute with a number to filter results. template: Return [#step1] [where] [#step2] [comparator] [number] \n Union: A union step is used to return results of two steps together. template: Return [#step1] [or / ,] [#step2]\n Intersection: An intersection step returns the result that two steps have in common. template: Return [attribute] of both [#step1] and [#step2]\n Discard: A discard step returns result of a step and excludes result of another step from it. template: Return [#step1] besides [#step2]\n Sort: A sort returns result of another step in a specific order. template: Return [#step1] [ordered / sorted by] [#step2]\n Is true: An is true step checks a condition on another result and returns a true or false. template: Return [is / if] [condition]\n Arithmetic: An arithmatic step operates an arithmatic operation on one or more steps. template: Return the [arithmetic op.] of [#step1] [and] [#step2].", | |
| "task1409-dart-text-generation": "In this task, you are given triplets. Each triplet is in the form of [subject, predicate, object]. Your task is to generate proper sentence that utilizes these triples. The objective is to construct a sentence that (a) captures the facts specified in the triples and (b) is a well-formed sentence easily understandable by a human. All triple values need not be used directly in the sentence as long as the facts are adequately captured.", | |
| "task421-persent-sentence-sentiment-classification": "Given a sentence and an entity, the task is to select the authors sentiment towards the enity. Sentiments can be Positive, Neutral and Negative. Select Positive if the sentence expresses a positive view towards the given entity or praises its quality or skills. Select Neutral if the sentence expresses no clear view towards the entity or has equal amounts of positive and negative statements or expressing some fact/quote by someone else. Select Negative if the sentence expresses a negative view towards like harsh remarks, criticizing entities action/decision etc. Note that URLs in the text have been replaced with [Link].", | |
| "task672-nummersense": "In this task, find the most appropriate number to replace the blank (indicated with _ ) and express it in words.", | |
| "task495-semeval-headline-classification": "Given news headlines and an edited word. The original sentence has word within given format {word}. Create new headlines by replacing {word} in the original sentence with edit word. Classify news headlines into \"Funny\" and \"Not Funny\" that have been modified by humans using an edit word to make them funny.", | |
| "task326-jigsaw-classification-obscene": "In this task, you are given a public comment from online platforms. You are expected to classify the comment into two classes: obscene and non-obscene. A comment is obscene if it is considered offensive to the public sense of decency, for example because it too obviously relates to sex or contains language regarded as taboo in polite usage.", | |
| "task311-race-question-generation": "In this task, you're given an article and an answer. Your task is to generate the question for the answer based on the given article.", | |
| "task1357-xlsum-summary-generation": "Generate an appropriate single-sentence summary for the given text such that it includes the main topic of the text.", | |
| "task493-review-polarity-classification": "Given reviews from Amazon, classify those review based on their content into two classes: Negative or Positive.", | |
| "task046-miscellaneous-question-typing": "You are given a question-answer pair. Answer with their type. Pay attention that there may be more than one correct type, but you only have to choose one. In your responses, use of the following types: \n(1) Humans: Any individual or group of humans, including fictional ones (e.g., a group or organization of persons , an individual, title of a person, description of a person); \n(2) Event: Any phenomenon natural or artificial (e.g., named hurricanes, Battles, Wars, Sports events, Terrorist attacks); \n(3) Entity: A thing with distinct and independent existence (Animals, Organs of body, Colors, Inventions, books and other creative pieces, Currency name, Diseases, and medicine, Food, Musical instrument, Languages, Plants, Products, Religions, Sports, Elements and substances, Symbols and signs, Techniques and methods, Equivalent terms, Vehicles); \n(4) Facility: Something built for a particular purpose (Buildings, Airports, Highways, Bridges); \n(5) Location: A place (Cities, Countries, Mountains, States); \n(6) Law: Named documents made into laws (e.g., “the first amendment”, \"civil rights act\"); \n(7) Organization: an organized body of people with a particular purpose (Company names, e.g. Google, Cults or terrorist groups, e.g. Al Qaeda); \n(8) Date: Absolute or relative dates or periods, bigger than 1 day (Years, Range, e.g. from Monday to Tuesday, or during the 20th century, Approximate time); \n(9) Time: Any temporal range/unit that is shorter than a day (e.g., 2 o'clock, 1 pm); \n(10) Money: Monetary values, including unit (e.g., \"$26\", \"914$\"); \n(11) Quantity: postcodes or other codes, the number of sth, Ranks, fractions, speed, temperature, size, area, and volume, weight (e.g., \"26 degree\" \"17 inch\"); \n(12) Description: description and abstract concepts (e.g., the definition of something, the manner of an action, reasons); \n(13) Abbreviation: expression abbreviated (e.g., AMT = abbreviation of Amazon Mechanical Turk). Don't generate any word that is not mentioned in the list of types (Humans, Event, Entity, Facility, Location, Law, Organization, Date, Time, Money, Quantity, Description, Abbreviation). If you can not associate any of the given types with the provided question and answer pair, respond \"Other\".", | |
| "task749-glucose-reverse-cause-emotion-detection": "In this task, you will be given a short story. One sentence from the story is chosen. Consider the likely emotions of the participants in the sentence and those affected by it. Is any of these emotions caused by the sentence? You should write your answer in the form \" A >Causes> B\". Try to use phrases and sentences from the story to compose your answer when possible. For the sentence describing the result, you must use the verb feel(s).", | |
| "task270-csrg-counterfactual-context-generation": "Given a premise, an initial context, an original ending, and a new ending, the task is to generate the counterfactual context that is aligned with the new ending. Each instance consists of a five-sentence story. The premise is the first sentence of a story, and the second sentence, which is the initial context, provides more information about the story's context. The original ending is the last three sentences of the story. Also, you are given a new ending that can provide the new story with the same premise. You should write a counterfactual context to explain the small differences between the original and new endings. More specifically, a story context contains the general plot of the story. And a counterfactual context is a slight modification to the initial context. To sum up, you should write the second sentence of a story based on the premise(first sentence) and the new ending(last three sentences) of the story.", | |
| "task1728-web-nlg-data-to-text": "You will be given one or more triples. The second part of each triple shows the relation between the first and the third element. Your task is to write a simple and short piece of text (sentence(s)) that describes the triples in natural language.", | |
| "task165-mcscript-question-answering-commonsense": "You are given a paragraph (Passage), a question (Question) and two answer options (Option1 and Option2). Your task is to choose the correct answer to the question based on the passage. Answer of the question cannot be found directly from the passage, it is indirectly indicated in the passage. Do not answer with words other than the two options, and do not answer with option titles(e.g. \"option1\" or \"1\" are not acceptable answers.)", | |
| "task510-reddit-tifu-title-summarization": "In this task, you are given a Reddit post as a text. Your task is to generate a title for this text. The title should start with \"TIFU by\", followed by a situation that caused humor. The title should contain 7-12 words, ideally.", | |
| "task566-circa-classification": "In this task, you are given two sentences. Your task is to classify the given sentences as \"Yes\" if they have same meaning; otherwise, classify them as \"No\".", | |
| "task223-quartz-explanation-generation": "In this task, you are given a question and answer for that. The answer will fill in the blank or will complete the unfinished question. Your task is to provide an explanation based on the given question or fill-in-the-blank statement, and answer.", | |
| "task222-rocstories-two-chioce-slotting-classification": "In this task, you're given the title and three arbitrary sentences out of a five-sentence story. You are also given three additional sentence options, a, b, and c, that may or may not belong to the story. Your job is to pick the two options that seamlessly connect with the rest of the story and their positions in the sentence order of the story; note that the selected choices may fit into the story at any point. Your answer must be in the form of '2a, 5b', where '2a' means the candidate sentence 'a' will be inserted as the 2nd sentence in the story. The answer must also be in the order of the selected choices, i.e., '2a, 5b' is allowed, and '5b, 2a' is not allowed. If options are equally plausible, pick the ones that make more sense.", | |
| "task231-iirc-link-classification": "In this task, you're given a question, a context passage, and four options which are terms from the passage. After reading a passage, you will get a brief understanding of the terms. Your job is to determine by searching and reading further information of which term you can answer the question. Indicate your choice as 'a', 'b', 'c', or 'd'. If you think more than one option is plausible, choose the more probable option to help you answer the question.", | |
| "task026-drop-question-generation": "This task involves creating questions from a given passage that involve some kind of complex reasoning (including numerical reasoning). \nThe generated questions must require looking at more than one part of the passage to answer. Try to use a variety of reasoning types in your questions (some of the sample reasoning types are illustrated in the 'positive examples' field) and also have a variety of answer types (spans, numbers, dates). A span is a continuous phrase taken directly from the passage or question. In such questions, the answer spans are recommended to be five words. Questions with answer type \"number\", are expected to start with \"How many\". For instance, \"How long did an event last?\" can be phrased as \"How many years did an event last?\". To increase diversity, try to create more questions for which the answer will be multiple spans. Please create questions that require AT LEAST 2 arithmetic operations as illustrated in the 'positive examples' field. You are encouraged to try DIFFERENT COMBINATIONS of operations and not just limit to the combinations shown in those examples. The more challenging questions you create, the better it is.", | |
| "task516-senteval-conjoints-inversion": "In this task you are given a sentence with one coordinating conjunction (for, and, nor, but, or, yet, and so). You must judge whether the order of two coordinated clausal conjoints have been inverted or not. Inversion would result in the sentence sounding unnatural. Label the instances as \"Original\" or \"Inversion\" based on your judgment.", | |
| "task1295-adversarial-qa-question-answering": "In this task, you are given a question and a context passage. You have to answer the question based on the given passage.", | |
| "task596-mocha-question-generation": "In this task, you're given a context passage. Your job is to generate relevant questions that can be answered by directly referring to the passage.", | |
| "task1297-qasc-question-answering": "In this task, you are given two facts, and a multiple-choice question. Based on the given facts, answer the question with index of the correct option (e.g, \"A\").", | |
| "task039-qasc-find-overlapping-words": "Generate an overlapping word between the given two sentences. When you find the overlapping words, they don't have to match exactly, e.g., \"survival\" and \"survive\" are valid overlapping words. Little words like \"the\" or \"of\" don't count! You must generate significant words which are not the stop words.", | |
| "task684-online-privacy-policy-text-information-type-generation": "Given a part of privacy policy text, identify the type of personal information which is collected, used, tracked or retained. The type of information should be present inside the given policy text, answer as 'Not Specified' otherwise", | |
| "task246-dream-question-generation": "In this task, you will be shown a conversation. You need to write a question with three choices for the conversation. Your question should be answerable based on the conversation and only have one correct answer. Note that you don't need to answer your question. \"W\" and \"M\" in the conversations stand for \"woman\" and \"man\".", | |
| "task162-count-words-starting-with-letter": "In this task, you need to count the number of words in a sentence that start with the given letter. Answer with numbers and not words.", | |
| "task154-tomqa-find-location-hard-noise": "Given a story, answer the question about the story. The question is the last sentence in the input. These stories can be difficult due to their length and how each story has at least one of the three following scenarios: the first is when the individual's belief matches reality, the second is when the individual's belief does not match reality, and the third is when an individual has a false belief about another individual's beliefs. The question will ask about the location of an object in the story with respect to either none or one of the three scenarios. Note that there are distractor sentences in each story that are unrelated to the question and are designed to confuse the reader.", | |
| "task1291-multi-news-summarization": "In this task, you are given a text of many news articles seperated by special token \"|||||\". Your task is to summarize them.", | |
| "task671-ambigqa-text-generation": "In this task, you're given an ambiguous question (which can be answered in more than one way). Your task is to provide one question which clarifies the input question and it has one unique answer, and also provide an answer to the generated question. Generated question and answer should be separated with a new line.", | |
| "task302-record-classification": "In this task, you will be presented with a passage and have to answer a question based on your understanding of the events inferred from the passage. Among the entities, try to find the best entity that is most likely to fill in \"_\" and classify the answers based on options.", | |
| "task1601-webquestions-answer-generation": "Based on the given question and tppic, give an answer. The answer is available on on the internet. The questions are mostly centered around a single named entity.", | |
| "task187-snli-entailment-to-contradiction-text-modification": "In this task, you're given a pair of sentences, sentence 1 and sentence 2, that agree with each other. Your job is to alter sentence 2 so that the pair contradict each other. Generated sentences must be short, with less than 15 words. New information can be introduced. Avoid using pronouns to confuse the subject of the sentence.", | |
| "task152-tomqa-find-location-easy-noise": "Given a story, answer the question about the story. The question is the last sentence in the input. The story has one of the three following scenarios: the first is when the individual's belief matches reality, the second is when the individual's belief does not match reality, and the third is when an individual has a false belief about another individual's beliefs. The question will ask about the location of an object in the story with respect to either none or one of the three scenarios. Note that there are distractor sentences in each story that are unrelated to the question and are designed to confuse the reader.", | |
| "task821-protoqa-question-generation": "Construct a question that every answer in the list is correct and reasonable for it.", | |
| "task1327-qa-zre-answer-generation-from-question": "Generate a correct and concise answer for the question based on the words in the context.", | |
| "task748-glucose-reverse-cause-event-detection": "In this task, you will be given a short story. One sentence from the story is chosen. Consider the events that happen after that sentence. Is any of them directly caused by it, or is made possible by it? You should write your answer in the form \" A >causes/enables> B\". Try to use phrases and sentences from the story to compose your answer when possible. Do not change the main selected sentence in your answer.", | |
| "task034-winogrande-question-modification-object": "In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, \"trophy\" and \"suitcase\". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from \"small\" to \"big\"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.\n| Attribute | triggerword | contrastive triggerword | \n| age | old | new | \n| altitude | low | high |\n| area | small | vast | \n| brightness | dark | light | \n| clarity | obscure | clear | \n| cleanness | dirty | clean | \n| complexity | simple | complex | \n| cost | cheap | expensive |\n| density | sparse | dense |\n| depth | shallow | deep |\n| distance | near | far | \n| electric conductivity | low | high |\n| flexibility | rigid | flexible |\n| granularity | fine | coarse | \n| hardness | soft | hard | \n| length | short | long | \n| magnitude | small | large | \n| mass | small | large | \n| odor | weak | strong | \n| pressure | low | high | \n| resistance | low | high | \n| shape | round | sharp |\n| shape | flat | spiky |\n| size | small | large | \n| sound | quiet | loud | \n| sound pitch | low | high | \n| speed | slow | fast | \n| stability | unstable | stable | \n| strength | weak | strong | \n| temperature | low | high | \n| texture | smooth | rough | \n| thermal conductivity | low | high |\n| thickness | thin | thick | \n| volume | small | large |\n| weight | light | heavy | \n| width | narrow | wide | \n| location | in | out |\n| location | up | down |\n| location | above | below |\n| location | on | off |\n| location | to | from |", | |
| "task023-cosmosqa-question-generation": "Based on the given context, craft a common-sense question, especially those that are LONG, INTERESTING, and COMPLEX. The goal is to write questions that are easy for humans and hard for AI machines! To create such questions, here are some suggestions: A. What may (or may not) be the plausible reason for an event? B. What may (or may not) happen before (or after, or during) an event? C. What may (or may not) be a plausible fact about someone (or something)? D. What may (or may not) happen if an event happens (or did not happen)? You can also create other types of questions. DO NOT make your question answerable without looking at the context, or question of which the correct answer can be directly extracted from the context. DO NOT ask a question that requires very specialized knowledge that is not common sense. DO NOT ask too simple or too short questions. Your question must be related to the context and answerable with common sense. Try to add more variations and complexity to the questions.", | |
| "task581-socialiqa-question-generation": "In this task, you're given context and an answer. Your task is to generate the question for this answer based on the given context with commonsense reasoning about social situations..", | |
| "task201-mnli-neutral-classification": "In this task, you're given a statement and three sentences as choices. Your job is to determine the neutral choice based on your inference from the statement and your commonsense knowledge. The neutral choice is a sentence that neither agrees nor disagrees with the statement. Indicate your answer as '1', '2', or '3', corresponding to the choice number of the selected sentence. If sentence X agrees with sentence Y, one's correctness follows from the other one. If sentence X disagrees with sentence Y, they can not be correct at the same time.", | |
| "task511-reddit-tifu-long-text-summarization": "In this task, you are given a Reddit post as a text. Your task is to generate a short summary for this text. The summary must include a situation which caused humor. The summary should be one or two sentences long." | |
| } |