id stringlengths 19 44 | messages listlengths 4 12 | state_supervision listlengths 1 5 | meta dict |
|---|---|---|---|
8072:early_fact_late_recall | [
{
"role": "user",
"content": "I'm planning a trip to Lake Baikal for the Baikal Ice Marathon. I have a severe gluten intolerance."
},
{
"role": "assistant",
"content": "That's important to know. Most events have gluten-free options at the support stations. What else would you like to know?"
},... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "8072:early_fact_late_recall:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X ... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "8072",
"scenario_id": "early_fact_late_recall",
"scenario_title": "Early Fact, Late Recall",
"scenario_instruction": "Introduce a concrete fact, label, preference, or commitment early, then make a later user turn depend on recalling it after at least one dis... |
5169:topic_pivot_then_return | [
{
"role": "user",
"content": "I'm trying to recall the author of \"Jane Eyre\" for a quiz."
},
{
"role": "assistant",
"content": "That was written by Charlotte Bronte."
},
{
"role": "user",
"content": "Quick question, what's the best way to roast a potato?"
},
{
"role": "assi... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5169:topic_pivot_then_return:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5169",
"scenario_id": "topic_pivot_then_return",
"scenario_title": "Topic Pivot Then Return",
"scenario_instruction": "Let the conversation briefly pivot to a secondary issue, then return to the original problem in a way that requires preserving the first th... |
1229:preserve_opening_retarget_later | [
{
"role": "user",
"content": "Identify the relationship between the words: Person - Job"
},
{
"role": "assistant",
"content": "That relationship indicates a person's profession."
},
{
"role": "user",
"content": "Can you list three people and their jobs?"
},
{
"role": "assista... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "1229:preserve_opening_retarget_later:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [Us... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "1229",
"scenario_id": "preserve_opening_retarget_later",
"scenario_title": "Preserve Opening, Retarget Later",
"scenario_instruction": "Keep the first user message exactly, but rewrite later turns so the conversation develops into a harder memory-dependent f... |
2862:same_surface_final_question_new_context | [
{
"role": "user",
"content": "We are launching a new organic waste collection service."
},
{
"role": "assistant",
"content": "That's interesting. Is it composting or recycling?"
},
{
"role": "user",
"content": "It's composting, but people call it recycling often."
},
{
"role"... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2862:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2862",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
2472:hard_topic_reset | [
{
"role": "user",
"content": "Describe a specific instance of bias in AI."
},
{
"role": "assistant",
"content": "One prominent example of bias in AI is facial recognition technology. This technology has been known to have higher accuracy when it comes to identifying people with lighter skin tone... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2472:hard_topic_reset:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [Assis... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2472",
"scenario_id": "hard_topic_reset",
"scenario_title": "Hard Topic Reset",
"scenario_instruction": "Make the user abruptly switch to a different topic for one turn, then later return to the original topic. The assistant must not let the temporary switch... |
7439:same_surface_final_question_new_context | [
{
"role": "user",
"content": "Make a speech about why consideration for the environment is important."
},
{
"role": "assistant",
"content": "Good morning everyone! Today, I'd like to talk about why consideration for the environment is so important. Every day, we are consuming vast amounts of the... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "7439:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "7439",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
5990:coreference_disambiguation | [
{
"role": "user",
"content": "Tell me about an allusion regarding a famous historical leader."
},
{
"role": "assistant",
"content": "There are several. George Washington is often called the Father of the Country, but Napoleon Bonaparte is also frequently referenced."
},
{
"role": "user",... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5990:coreference_disambiguation:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5990",
"scenario_id": "coreference_disambiguation",
"scenario_title": "Coreference Disambiguation",
"scenario_instruction": "Introduce at least two plausible referents of the same type, then make a later user follow-up rely on a pronoun or shorthand whose me... |
4135:two_thread_tracking | [
{
"role": "user",
"content": "Write a C# method to calculate compound interest given principal, rate, and periods."
},
{
"role": "assistant",
"content": "Here is the C# method. It takes the principal, rate, and periods as arguments and calculates the total amount using the compound interest form... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "4135:two_thread_tracking:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "4135",
"scenario_id": "two_thread_tracking",
"scenario_title": "Two Thread Tracking",
"scenario_instruction": "Create two interleaved subproblems inside one conversation and ensure the assistant must keep both threads straight rather than collapsing them tog... |
5109:clarify_then_resolve | [
{
"role": "user",
"content": "I need to retrieve a word we discussed, but I'm not certain if the capitalization matters for my search query."
},
{
"role": "assistant",
"content": "I recall storing 'elephant' and 'ELEPHANT'. Would you prefer me to treat them as identical or distinct for your sear... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5109:clarify_then_resolve:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [A... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5109",
"scenario_id": "clarify_then_resolve",
"scenario_title": "Clarify Then Resolve",
"scenario_instruction": "Make the user start vague, then reveal a crucial detail only after clarification. The final answer should depend on the revealed detail rather th... |
3226:interrupting_followup_chain | [
{
"role": "user",
"content": "Translate \"I am happy\" into Spanish."
},
{
"role": "assistant",
"content": "Estoy feliz."
},
{
"role": "user",
"content": "Wait, what is \"thank you\" in French?"
},
{
"role": "assistant",
"content": "The phrase is \"merci\"."
},
{
... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "3226:interrupting_followup_chain:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "3226",
"scenario_id": "interrupting_followup_chain",
"scenario_title": "Interrupting Follow-up Chain",
"scenario_instruction": "Break an otherwise smooth line of reasoning with a short interruption or side question, then resume the original thread cleanly.",... |
4298:user_attention_test | [
{
"role": "user",
"content": "Change the given set of words into a complete sentence: the growth of online shopping"
},
{
"role": "assistant",
"content": "The growth of online shopping has been increasing significantly. According to a report by Nielsen, e-commerce penetration in the US jumped fr... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "4298:user_attention_test:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "4298",
"scenario_id": "user_attention_test",
"scenario_title": "User Attention Test",
"scenario_instruction": "Make the user explicitly test whether the assistant was paying attention to an earlier detail or instruction.",
"must_have": [
"The user must... |
5490:late_request_for_summary_of_prior_state | [
{
"role": "user",
"content": "Calculate the trade-off cost between Investment 1 ($100) and Investment 2 ($80)."
},
{
"role": "assistant",
"content": "To calculate the cost trade-off, I need to know the potential returns and risk levels for each investment."
},
{
"role": "user",
"cont... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5490:late_request_for_summary_of_prior_state:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5490",
"scenario_id": "late_request_for_summary_of_prior_state",
"scenario_title": "Late Summary of Prior State",
"scenario_instruction": "Have the later user ask for a compact summary, recap, or decision based on several earlier details that were introduced... |
84:user_attention_test | [
{
"role": "user",
"content": "Convert 1010 from binary to decimal. Constraint: Do not use a period at the end."
},
{
"role": "assistant",
"content": "1010 in binary is 10"
},
{
"role": "user",
"content": "Okay, I want to continue. Convert 1100 from binary to decimal. Keep the no-peri... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "84:user_attention_test:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [Assi... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "84",
"scenario_id": "user_attention_test",
"scenario_title": "User Attention Test",
"scenario_instruction": "Make the user explicitly test whether the assistant was paying attention to an earlier detail or instruction.",
"must_have": [
"The user must t... |
9320:interrupting_followup_chain | [
{
"role": "user",
"content": "How can schools best prepare students for the workforce?"
},
{
"role": "assistant",
"content": "Schools can best prepare students for the workforce by providing hands-on job experience or internships, giving them the opportunity to learn soft skills such as communic... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "9320:interrupting_followup_chain:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "9320",
"scenario_id": "interrupting_followup_chain",
"scenario_title": "Interrupting Follow-up Chain",
"scenario_instruction": "Break an otherwise smooth line of reasoning with a short interruption or side question, then resume the original thread cleanly.",... |
7362:early_fact_late_recall | [
{
"role": "user",
"content": "Plot the 2nd column against the 1st column of the data: 1 4 7 2 5 8 3 6 9. Please set the X-axis label to 'Quantity' and the Y-axis label to 'Price'."
},
{
"role": "assistant",
"content": "Here is the matplotlib code to plot your data with the requested labels. ```p... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "7362:early_fact_late_recall:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X ... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "7362",
"scenario_id": "early_fact_late_recall",
"scenario_title": "Early Fact, Late Recall",
"scenario_instruction": "Introduce a concrete fact, label, preference, or commitment early, then make a later user turn depend on recalling it after at least one dis... |
4098:late_request_for_summary_of_prior_state | [
{
"role": "user",
"content": "Describe how the COVID-19 virus spreads."
},
{
"role": "assistant",
"content": "The COVID-19 virus is spread through contact with contaminated surfaces or through the air from respiratory droplets when an infected person coughs, sneezes, or talks. It can also be spr... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "4098:late_request_for_summary_of_prior_state:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "4098",
"scenario_id": "late_request_for_summary_of_prior_state",
"scenario_title": "Late Summary of Prior State",
"scenario_instruction": "Have the later user ask for a compact summary, recap, or decision based on several earlier details that were introduced... |
9475:late_constraint_reversal | [
{
"role": "user",
"content": "Convert the following sentence into present perfect tense: 'He spoke at the conference.'"
},
{
"role": "assistant",
"content": "He has spoken at the conference."
},
{
"role": "user",
"content": "Do the same for: 'She wrote a book.'"
},
{
"role": ... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "9475:late_constraint_reversal:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: ... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "9475",
"scenario_id": "late_constraint_reversal",
"scenario_title": "Late Constraint Reversal",
"scenario_instruction": "Make the assistant start solving one version of the problem, then add a late user constraint that forces the assistant to adapt without f... |
9156:two_thread_tracking | [
{
"role": "user",
"content": "Can you list cat breeds that have distinct stripes?"
},
{
"role": "assistant",
"content": "Sure, the Tabby is the most common striped breed, and Bengals also have distinct markings."
},
{
"role": "user",
"content": "That's helpful. For a small apartment,... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "9156:two_thread_tracking:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "9156",
"scenario_id": "two_thread_tracking",
"scenario_title": "Two Thread Tracking",
"scenario_instruction": "Create two interleaved subproblems inside one conversation and ensure the assistant must keep both threads straight rather than collapsing them tog... |
9523:user_attention_test | [
{
"role": "user",
"content": "Describe how a business could use chatbots to improve customer experience. I am a hospital administrator."
},
{
"role": "assistant",
"content": "Chatbots can improve customer experience through 24/7 support. I will tailor this for a hospital administrator."
},
{... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "9523:user_attention_test:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "9523",
"scenario_id": "user_attention_test",
"scenario_title": "User Attention Test",
"scenario_instruction": "Make the user explicitly test whether the assistant was paying attention to an earlier detail or instruction.",
"must_have": [
"The user must... |
984:same_surface_final_question_new_context | [
{
"role": "user",
"content": "I'm working on a system to transcribe medical doctor-patient calls."
},
{
"role": "assistant",
"content": "You'll first need an Automatic Speech Recognition (ASR) pipeline to convert audio to text."
},
{
"role": "user",
"content": "Do the embeddings need... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "984:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following form... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "984",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to ... |
9843:coreference_disambiguation | [
{
"role": "user",
"content": "In the sentence 'She worked diligently', is 'diligently' an adjective or adverb?"
},
{
"role": "assistant",
"content": "It is an adverb. It modifies the verb 'worked'."
},
{
"role": "user",
"content": "Okay. And in 'She worked hard', is 'hard' an adjecti... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "9843:coreference_disambiguation:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "9843",
"scenario_id": "coreference_disambiguation",
"scenario_title": "Coreference Disambiguation",
"scenario_instruction": "Introduce at least two plausible referents of the same type, then make a later user follow-up rely on a pronoun or shorthand whose me... |
1728:two_thread_tracking | [
{
"role": "user",
"content": "Construct a proverb that encapsulates the concept of freedom."
},
{
"role": "assistant",
"content": "Freedom lays the foundation for a life of true happiness."
},
{
"role": "user",
"content": "Can you give me an example of a proverb that reflects the imp... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "1728:two_thread_tracking:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "1728",
"scenario_id": "two_thread_tracking",
"scenario_title": "Two Thread Tracking",
"scenario_instruction": "Create two interleaved subproblems inside one conversation and ensure the assistant must keep both threads straight rather than collapsing them tog... |
5440:user_attention_test | [
{
"role": "user",
"content": "Provide three possible solutions to reduce plastic waste."
},
{
"role": "assistant",
"content": "1. Reducing single-use plastic by using reusable bags. 2. Using plastic-to-fuel technology. 3. Plastic-eating bacteria."
},
{
"role": "user",
"content": "Can... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5440:user_attention_test:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5440",
"scenario_id": "user_attention_test",
"scenario_title": "User Attention Test",
"scenario_instruction": "Make the user explicitly test whether the assistant was paying attention to an earlier detail or instruction.",
"must_have": [
"The user must... |
1290:early_fact_late_recall | [
{
"role": "user",
"content": "Write a metaphor for exhaustion."
},
{
"role": "assistant",
"content": "It was like a car running on empty."
},
{
"role": "user",
"content": "Cool, what about sleepiness?"
},
{
"role": "assistant",
"content": "It was like a fog."
},
{
... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "1290:early_fact_late_recall:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X ... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "1290",
"scenario_id": "early_fact_late_recall",
"scenario_title": "Early Fact, Late Recall",
"scenario_instruction": "Introduce a concrete fact, label, preference, or commitment early, then make a later user turn depend on recalling it after at least one dis... |
2433:early_fact_late_recall | [
{
"role": "user",
"content": "Remember, I only want examples that start with a verb."
},
{
"role": "assistant",
"content": "Understood, I will follow that."
},
{
"role": "user",
"content": "Okay, give me an interrogative sentence."
},
{
"role": "assistant",
"content": "Is... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2433:early_fact_late_recall:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X ... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2433",
"scenario_id": "early_fact_late_recall",
"scenario_title": "Early Fact, Late Recall",
"scenario_instruction": "Introduce a concrete fact, label, preference, or commitment early, then make a later user turn depend on recalling it after at least one dis... |
1541:late_constraint_reversal | [
{
"role": "user",
"content": "What are the steps to weeding a garden?"
},
{
"role": "assistant",
"content": "The steps to weeding a garden include: 1. Prepare the soil: Loosen the soil around the weeds and remove any debris. 2. Pull the weeds: Pull the unwanted plants, making sure to get the ent... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "1541:late_constraint_reversal:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: ... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "1541",
"scenario_id": "late_constraint_reversal",
"scenario_title": "Late Constraint Reversal",
"scenario_instruction": "Make the assistant start solving one version of the problem, then add a late user constraint that forces the assistant to adapt without f... |
5039:reference_previous_assistant_statement | [
{
"role": "user",
"content": "Generate a list of ten famous monuments in the world."
},
{
"role": "assistant",
"content": "1. Taj Mahal, India 2. Great Wall of China 3. Eiffel Tower, France 4. Christ the Redeemer, Brazil 5. Colosseum, Italy 6. Petra, Jordan 7. Statue of Liberty, USA 8. Angkor Wa... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5039:reference_previous_assistant_statement:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following form... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5039",
"scenario_id": "reference_previous_assistant_statement",
"scenario_title": "Reference Previous Assistant Statement",
"scenario_instruction": "Make a later user turn refer back to something the assistant itself said earlier, requiring the assistant to ... |
824:long_explanation_then_specific_followup | [
{
"role": "user",
"content": "What are the primary security risks regarding online banking?"
},
{
"role": "assistant",
"content": "The main risks are identity theft, malware, and financial fraud."
},
{
"role": "user",
"content": "You mentioned malware. Could you explain the specific ... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "824:long_explanation_then_specific_followup:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following form... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "824",
"scenario_id": "long_explanation_then_specific_followup",
"scenario_title": "Long Explanation Then Specific Follow-up",
"scenario_instruction": "Make the assistant provide a fairly detailed middle-turn explanation, then have the user ask a specific fol... |
2042:same_surface_final_question_new_context | [
{
"role": "user",
"content": "I want to focus on the top rising stars in the soccer world."
},
{
"role": "assistant",
"content": "Certainly. Kylian Mbappe is currently the most prominent rising star in soccer."
},
{
"role": "user",
"content": "Are there other notable names to conside... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2042:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2042",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
7460:same_surface_final_question_new_context | [
{
"role": "user",
"content": "What is the main idea of this text?: The robotics industry has become one of the fastest growing sectors in today's economy. Robotics technology is being applied in various industries and applications, such as manufacturing, healthcare, and defense."
},
{
"role": "assis... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "7460:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "7460",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
3220:same_surface_final_question_new_context | [
{
"role": "user",
"content": "What were the economic criticisms of the New Deal policies?"
},
{
"role": "assistant",
"content": "Critics of the New Deal argued that while it expanded federal power to address the Great Depression, it still operated within a capitalist framework. Opponents felt it... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "3220:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "3220",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
7090:user_attention_test | [
{
"role": "user",
"content": "List five uses for a paperclip, but do not include any items used for hanging."
},
{
"role": "assistant",
"content": "1. Binder clip. 2. Cable tie. 3. Flag. 4. Pen holder. 5. Hair tie."
},
{
"role": "user",
"content": "Now list five uses for a stapler."
... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "7090:user_attention_test:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "7090",
"scenario_id": "user_attention_test",
"scenario_title": "User Attention Test",
"scenario_instruction": "Make the user explicitly test whether the assistant was paying attention to an earlier detail or instruction.",
"must_have": [
"The user must... |
2023:long_explanation_then_specific_followup | [
{
"role": "user",
"content": "Convert 2F35 Hex to binary."
},
{
"role": "assistant",
"content": "The binary is 0010 1111 0011 0101. I calculated this by converting each hex digit to its 4-bit binary equivalent: 2 maps to 0010, F maps to 1111, 3 maps to 0011, and 5 maps to 0101."
},
{
"ro... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2023:long_explanation_then_specific_followup:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2023",
"scenario_id": "long_explanation_then_specific_followup",
"scenario_title": "Long Explanation Then Specific Follow-up",
"scenario_instruction": "Make the assistant provide a fairly detailed middle-turn explanation, then have the user ask a specific fo... |
2770:hard_topic_reset | [
{
"role": "user",
"content": "Troubleshoot the following error message: \"An unknown error occurred. Error code: 500.\""
},
{
"role": "assistant",
"content": "The error code 500 is an internal server error, indicating that something has gone wrong on the server side of the application. It could ... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2770:hard_topic_reset:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [Assis... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2770",
"scenario_id": "hard_topic_reset",
"scenario_title": "Hard Topic Reset",
"scenario_instruction": "Make the user abruptly switch to a different topic for one turn, then later return to the original topic. The assistant must not let the temporary switch... |
810:preserve_opening_retarget_later | [
{
"role": "user",
"content": "List the 3 longest rivers in Europe."
},
{
"role": "assistant",
"content": "The three longest rivers in Europe are the Volga, the Danube, and the Ural. Note that the Dnieper is the fourth-longest."
},
{
"role": "user",
"content": "You mentioned the Volga... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "810:preserve_opening_retarget_later:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [Use... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "810",
"scenario_id": "preserve_opening_retarget_later",
"scenario_title": "Preserve Opening, Retarget Later",
"scenario_instruction": "Keep the first user message exactly, but rewrite later turns so the conversation develops into a harder memory-dependent fo... |
5847:topic_pivot_then_return | [
{
"role": "user",
"content": "Using the given information, construct a paragraph about the history of the company: Company Name: Heavy Machinery Year Established: 1982"
},
{
"role": "assistant",
"content": "Heavy Machinery was founded in 1982 by John Johnson, an experienced engineer from the nea... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5847:topic_pivot_then_return:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5847",
"scenario_id": "topic_pivot_then_return",
"scenario_title": "Topic Pivot Then Return",
"scenario_instruction": "Let the conversation briefly pivot to a secondary issue, then return to the original problem in a way that requires preserving the first th... |
6785:two_thread_tracking | [
{
"role": "user",
"content": "I want to train a GPT model to perform simple addition and subtraction like 8 - 4. What dataset format do I need?"
},
{
"role": "assistant",
"content": "For arithmetic, you need a dataset with pairs of numbers and their corresponding results (e.g., 8 - 4 and 4)."
... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "6785:two_thread_tracking:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [As... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "6785",
"scenario_id": "two_thread_tracking",
"scenario_title": "Two Thread Tracking",
"scenario_instruction": "Create two interleaved subproblems inside one conversation and ensure the assistant must keep both threads straight rather than collapsing them tog... |
9436:same_surface_final_question_new_context | [
{
"role": "user",
"content": "Tell me a pun about 'lion'."
},
{
"role": "assistant",
"content": "A lion is a king of the jungle."
},
{
"role": "user",
"content": "Tell me a pun about 'tiger'."
},
{
"role": "assistant",
"content": "A tiger is a king of the jungle."
},
... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "9436:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "9436",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
2320:hard_topic_reset | [
{
"role": "user",
"content": "Find the length of the hypotenuse given two sides: side_length1 = 5, side_length2 = 6"
},
{
"role": "assistant",
"content": "The length of the hypotenuse is 7.81."
},
{
"role": "user",
"content": "What's the capital of France?"
},
{
"role": "assi... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "2320:hard_topic_reset:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]: X [Assis... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "2320",
"scenario_id": "hard_topic_reset",
"scenario_title": "Hard Topic Reset",
"scenario_instruction": "Make the user abruptly switch to a different topic for one turn, then later return to the original topic. The assistant must not let the temporary switch... |
7762:late_request_for_summary_of_prior_state | [
{
"role": "user",
"content": "Calculate the area of a triangle with a base of 5cm and a height of 6cm."
},
{
"role": "assistant",
"content": "The area of the triangle is 15 square centimeters, calculated using half the base times the height."
},
{
"role": "user",
"content": "What abo... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "7762:late_request_for_summary_of_prior_state:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "7762",
"scenario_id": "late_request_for_summary_of_prior_state",
"scenario_title": "Late Summary of Prior State",
"scenario_instruction": "Have the later user ask for a compact summary, recap, or decision based on several earlier details that were introduced... |
1052:deferred_question_then_answer | [
{
"role": "user",
"content": "Reverse engineer the following code to make the output \"odd\": if (x % 2 == 0) { print(\"even\") }"
},
{
"role": "assistant",
"content": "Here is the modified code: if (x % 2 != 0) { print(\"odd\") }. I'll explain the logic later."
},
{
"role": "user",
... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "1052:deferred_question_then_answer:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "1052",
"scenario_id": "deferred_question_then_answer",
"scenario_title": "Deferred Question",
"scenario_instruction": "Have the user ask something important early but explicitly postpone the answer while discussing something else, then request the answer lat... |
8448:same_surface_final_question_new_context | [
{
"role": "user",
"content": "I am writing a thank you letter to my professor for his mentorship and found this stanza: 'He is the only one Too much to him I owe His grace more sweet than honey'."
},
{
"role": "assistant",
"content": "That captures the sentiment well. For a title, how about 'Deb... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "8448:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "8448",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
338:coreference_disambiguation | [
{
"role": "user",
"content": "Give me two similes about \"light\"."
},
{
"role": "assistant",
"content": "Light is like a flood. Light is like a flame."
},
{
"role": "user",
"content": "Which one feels more dangerous?"
},
{
"role": "assistant",
"content": "The flame. It f... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "338:coreference_disambiguation:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User 1]:... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "338",
"scenario_id": "coreference_disambiguation",
"scenario_title": "Coreference Disambiguation",
"scenario_instruction": "Introduce at least two plausible referents of the same type, then make a later user follow-up rely on a pronoun or shorthand whose mea... |
5698:deferred_question_then_answer | [
{
"role": "user",
"content": "Name three countries with tropical climates."
},
{
"role": "assistant",
"content": "I can help. But before I list them, do you have a specific budget in mind?"
},
{
"role": "user",
"content": "I'm looking for affordable options."
},
{
"role": "as... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "5698:deferred_question_then_answer:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following format: [User... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "5698",
"scenario_id": "deferred_question_then_answer",
"scenario_title": "Deferred Question",
"scenario_instruction": "Have the user ask something important early but explicitly postpone the answer while discussing something else, then request the answer lat... |
7781:same_surface_final_question_new_context | [
{
"role": "user",
"content": "We are building a campaign for a new AI diagnostic platform. What is the marketing funnel?"
},
{
"role": "assistant",
"content": "The marketing funnel is a customer journey, starting with the moment a customer first interacts with the brand, then progressing through... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "7781:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "7781",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
1808:same_surface_final_question_new_context | [
{
"role": "user",
"content": "Select all the words that start with a capital letter from the sentence below: This is a Sentence with Several Capitals."
},
{
"role": "assistant",
"content": "This, Sentence, Several, Capitals."
},
{
"role": "user",
"content": "Can you provide another s... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "1808:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "1808",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
4513:same_surface_final_question_new_context | [
{
"role": "user",
"content": "Give me a synonym for the word 'love'."
},
{
"role": "assistant",
"content": "Affection."
},
{
"role": "user",
"content": "Can you give me an example sentence using the word 'affection' that involves a sibling relationship?"
},
{
"role": "assista... | [
{
"prefix_pair_index": 1,
"probes": [
{
"id": "4513:same_surface_final_question_new_context:transcript_probe:p1",
"family": "canonical_transcript_verbatim",
"question": "What is everything we have talked about so far? Give exact conversation transcript verbatim in following for... | {
"source_dataset": "ChatAlpaca",
"source_row_id": "4513",
"scenario_id": "same_surface_final_question_new_context",
"scenario_title": "Same Surface Final Question, New Context",
"scenario_instruction": "Reshape the earlier history so the final user message keeps nearly the same surface form but now refers to... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.