answer stringlengths 2 697 | deep_reasoning stringlengths 175 534 | dialogue_history listlengths 3 65 | dialogue_id stringlengths 21 51 | domain stringclasses 14
values | id stringlengths 41 72 | is_high_quality bool 2
classes | metadata unknown | quality_flags listlengths 0 10 | quality_score float64 0 1 | source_dataset stringclasses 3
values | source_id stringlengths 11 40 | split stringclasses 1
value | streaming_chunks listlengths 3 65 | streaming_reasoning listlengths 3 65 | task_type stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
No, thanks. | The dialogue is uncertain context in general daily life, with information request intent. The most recent user context is: Certainly. Would you like it with ice cream? The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "Can I help you?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, I'm in a bit of a rush. Could I have a piece of apple pie please?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Certainly. Would you like it with ice cream?",
"turn_i... | DailyDialog_test_000093 | daily_advice | life_mt_sct_DailyDialog_test_000093_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"general_daily_life"
],
"domain_catego... | [
"fragment_answer",
"low_information_answer",
"low_taxonomy_confidence",
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0.17 | DailyDialog | test_000093 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
There is a train TR0798 leaving at 9:29 and arriving at 10:07, would this time work for you? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=tonight, thursday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Hi there. I'm trying to find a place to eat tonight. I'd like it to be more high end and expensive, and serve gastropub food. Can you help me?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I have three Gastropub restaurants that meet your criteria. Two ar... | MultiWOZ_MUL1604.json | travel | life_mt_sct_MultiWOZ_MUL1604.json_target8 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"weak_slot_evidence"
] | 0.749 | MultiWOZ | MUL1604.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | prefix_to_next_assistant_text |
Do you have any other preferences? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=4 days; origin=la; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "I need a flight from LA to Miami.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I can help you with that. Can you please confirm the dates of your travel?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to leave today latest time po... | Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | travel | life_mt_sct_Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2_target8 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"weak_slot_evidence"
] | 0.749 | Taskmaster | dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
I'm afraid he's fully engaged.Could you make it another time? | The dialogue is schedule planning in personal schedule, with planning coordination intent. Key constraints are date_or_time=monday, afternoon; event_type=appointment, meeting. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "I'd like to make an appointment with Dr.Cooper.Could you arrange it for me?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes.Would tomorrow morning be all right with you?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Sorry, I've got a m... | DailyDialog_test_000094 | schedule | life_mt_sct_DailyDialog_test_000094_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"personal_schedule"
],
"domain_categor... | [] | 1 | DailyDialog | test_000094 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: event_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium"... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: event_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium"... | prefix_to_next_assistant_text |
It will take 38 minutes and cost 10.10 pounds. | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=tonight, thursday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Hi there. I'm trying to find a place to eat tonight. I'd like it to be more high end and expensive, and serve gastropub food. Can you help me?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I have three Gastropub restaurants that meet your criteria. Two ar... | MultiWOZ_MUL1604.json | travel | life_mt_sct_MultiWOZ_MUL1604.json_target10 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"premature_respond",
"weak_slot_evidence"
] | 0.749 | MultiWOZ | MUL1604.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | prefix_to_next_assistant_text |
Okay, let me find you something. I have found a non-stop flight, departing Los Angeles at 11:15 PM, arrives in Miami at 7:15 AM. It's an overnight flight. The cost is $2390. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=4 days; origin=la; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "I need a flight from LA to Miami.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I can help you with that. Can you please confirm the dates of your travel?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to leave today latest time po... | Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | travel | life_mt_sct_Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2_target10 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"premature_respond",
"weak_slot_evidence"
] | 0.749 | Taskmaster | dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
I have you booked for one seat on TR0798 departing Bishops Stortford for Cambridge at 05:98 and arriving at 06:07. Reference is M377ZH73. Anything else I can help with? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=tonight, thursday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Hi there. I'm trying to find a place to eat tonight. I'd like it to be more high end and expensive, and serve gastropub food. Can you help me?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I have three Gastropub restaurants that meet your criteria. Two ar... | MultiWOZ_MUL1604.json | travel | life_mt_sct_MultiWOZ_MUL1604.json_target12 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"weak_slot_evidence"
] | 0.749 | MultiWOZ | MUL1604.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | prefix_to_next_assistant_text |
Let me get your return flight details. I have found a non-stop flight departing Miami at 6:55 AM arriving Los Angeles at 10:20 AM. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=4 days; origin=la; transport_type=flight. The answer should address the customer support step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "I need a flight from LA to Miami.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I can help you with that. Can you please confirm the dates of your travel?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to leave today latest time po... | Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | travel | life_mt_sct_Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2_target12 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"premature_respond",
"weak_slot_evidence"
] | 0.749 | Taskmaster | dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
Thank you for contacting us. Enjoy your visit and have a great day. | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=tonight, thursday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Hi there. I'm trying to find a place to eat tonight. I'd like it to be more high end and expensive, and serve gastropub food. Can you help me?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I have three Gastropub restaurants that meet your criteria. Two ar... | MultiWOZ_MUL1604.json | travel | life_mt_sct_MultiWOZ_MUL1604.json_target14 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"premature_respond",
"weak_slot_evidence"
] | 0.749 | MultiWOZ | MUL1604.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: date_or_duration; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence"... | prefix_to_next_assistant_text |
Would you like me to book it for you? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=4 days; origin=la; transport_type=flight. The answer should address the confirmation clarification step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "I need a flight from LA to Miami.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I can help you with that. Can you please confirm the dates of your travel?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to leave today latest time po... | Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | travel | life_mt_sct_Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2_target14 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"weak_slot_evidence"
] | 0.749 | Taskmaster | dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
The booking has been done and the details have been emailed to you. Enjoy your trip. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=4 days; origin=la; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "I need a flight from LA to Miami.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I can help you with that. Can you please confirm the dates of your travel?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to leave today latest time po... | Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | travel | life_mt_sct_Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2_target16 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"premature_respond",
"weak_slot_evidence"
] | 0.749 | Taskmaster | dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
You're welcome. Goodbye. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=4 days; origin=la; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "I need a flight from LA to Miami.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "I can help you with that. Can you please confirm the dates of your travel?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to leave today latest time po... | Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | travel | life_mt_sct_Taskmaster_dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2_target18 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"fragment_answer",
"low_information_answer",
"too_short_answer",
"weak_slot_evidence",
"weak_target_answer"
] | 0.15 | Taskmaster | dlg-02d3cc76-c1bb-4b86-9c08-48b01dbeaaf2 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
Okay, to where? | The visible history is a flight booking task in travel transportation. It should track origin, destination, and travel dates. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target4 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0.65 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
No, I haven't got one by now. I will go to the interview of yours. | The dialogue is job-interview in education career, with information request intent. The most recent user context is: I want to know whether you will come to the interview. So have you accepted offers from other companies? The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "Hello, is that Xiaogang Li who applied for the Solar Studio of our college?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, speaking.",
"turn_id": 2
},
{
"speaker": "user",
"text": "I want to know whether you will come to the interview.... | DailyDialog_test_000096 | other_life | life_mt_sct_DailyDialog_test_000096_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"education_career"
],
"domain_category... | [
"weak_final_state"
] | 0.97 | DailyDialog | test_000096 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "high",
"r... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "high",
"r... | prefix_to_next_assistant_text |
Okay. And what else? | The visible history is a flight booking task in travel transportation. It should track origin, destination, and travel dates. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target6 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"weak_final_state"
] | 0.899 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Let's make sure we are on the same page. Do we want a formal or informal gathering? | The dialogue is workplace conversation in social relationship, with information request intent. The most recent user context is: I have no idea what the entertainment should be. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "I need help with the office party.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, of course. We could split it. What part would you like to do, the food or the entertainment?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I have no id... | DailyDialog_test_000097 | social_planning | life_mt_sct_DailyDialog_test_000097_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"weak_final_state"
] | 0.97 | DailyDialog | test_000097 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Okay. Anything else? | The visible history is a flight booking task in travel transportation. It should track origin, destination, and travel dates. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target8 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0.65 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Yes, that would be best. Now, for food should we have Chinese or American food? | The dialogue is workplace conversation in social relationship, with recommendation intent. The most recent user context is: Let's make it a formal affair. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "I need help with the office party.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, of course. We could split it. What part would you like to do, the food or the entertainment?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I have no id... | DailyDialog_test_000097 | social_planning | life_mt_sct_DailyDialog_test_000097_target6 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"weak_final_state"
] | 0.97 | DailyDialog | test_000097 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Okay. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target10 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"answer_not_grounded",
"fragment_answer",
"generic_answer",
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
That would be great. What kind of music would you like, a band or recorded music? | The dialogue is workplace conversation in social relationship, with recommendation intent. The most recent user context is: American food would be the best choice. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "I need help with the office party.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, of course. We could split it. What part would you like to do, the food or the entertainment?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I have no id... | DailyDialog_test_000097 | social_planning | life_mt_sct_DailyDialog_test_000097_target8 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"weak_final_state"
] | 0.97 | DailyDialog | test_000097 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Unfortunately there are not guesthouses that meet your criteria. Would you like to look for something in the moderate price range or a hotel instead? | The visible history is a hotel search task in travel transportation. It should track area, price, and amenities. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target4 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"weak_final_state"
] | 0.97 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Sure thing, first class starts at $10000. Is that okay? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target12 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"weak_final_state"
] | 0.97 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
That works for me. Let's see what we can each do about planning our part and meet again on Friday. | The dialogue is workplace conversation in social relationship, with information request intent. The most recent user context is: We should bring in a really good DJ. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "I need help with the office party.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, of course. We could split it. What part would you like to do, the food or the entertainment?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I have no id... | DailyDialog_test_000097 | social_planning | life_mt_sct_DailyDialog_test_000097_target10 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"weak_final_state"
] | 0.97 | DailyDialog | test_000097 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
I do not have any expensive guesthouses with wifi and parking. I do have some moderate or cheap ones. | The visible history is a hotel search task in travel transportation. It should track area, price, and amenities. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target6 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"premature_respond",
"weak_final_state"
] | 0.92 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay. What else? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target14 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0.65 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
No, thank you. | The dialogue is uncertain context in general daily life, with information request intent. The most recent user context is: Would you care for some lemon for your tea? The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "What would you like to eat, sir?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Scrambled egg, bacon, three pieces of bread and a cup of tea.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Would you care for some lemon for your tea?",
... | DailyDialog_test_000099 | cooking | life_mt_sct_DailyDialog_test_000099_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"general_daily_life"
],
"domain_catego... | [
"fragment_answer",
"low_information_answer",
"low_taxonomy_confidence",
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0.17 | DailyDialog | test_000099 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
I didn't find any that matched that criteria. | The visible history is a hotel search task in travel transportation. Known constraints are hotel_preference=hotel. The answer should address the recommendation step while tracking area, price, and amenities. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target8 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 0.899 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay. Anything else? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target16 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"target_leakage",
"too_short_answer",
"weak_final_state",
"weak_target_answer"
] | 0.499 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
I bid on a Chanel watch at the last minute, and got it sixty percent off! | The dialogue is uncertain context in general daily life, with small talk intent. The most recent user context is: But there are only two more days until the bidding closes! The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "Only one. But I wanted to make sure I'd get it.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yeah, I really wanted that Chanel bag, too. But just in case you should bid low.",
"turn_id": 2
},
{
"speaker": "user",
"text": "But there are on... | DailyDialog_test_000101 | other_life | life_mt_sct_DailyDialog_test_000101_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"general_daily_life"
],
"domain_catego... | [
"low_taxonomy_confidence",
"weak_final_state"
] | 0.749 | DailyDialog | test_000101 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
I have a few hotels that meet your needs. How about the Gonville Hotel? It is in the centre of town. | The visible history is a hotel search task in travel transportation. Known constraints are hotel_preference=hotel. The answer should address the information request step while tracking area, price, and amenities. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target10 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 1 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
When did you want to fly? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target18 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"weak_final_state"
] | 0.899 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Don't count on it, Babe. eBay is always full of surprises. | The dialogue is uncertain context in general daily life, with small talk intent. The most recent user context is: Oh... I'm such a newbie. But at least I'm sure I'll get that tennis racket. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "Only one. But I wanted to make sure I'd get it.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yeah, I really wanted that Chanel bag, too. But just in case you should bid low.",
"turn_id": 2
},
{
"speaker": "user",
"text": "But there are on... | DailyDialog_test_000101 | other_life | life_mt_sct_DailyDialog_test_000101_target6 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"general_daily_life"
],
"domain_catego... | [
"answer_not_grounded",
"low_taxonomy_confidence",
"weak_final_state"
] | 0.499 | DailyDialog | test_000101 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
I have booked you at the gonville hotel for 1 night however we will need to know for how long and how many people. Thank you. | The visible history is a hotel booking task in travel transportation. Known constraints are hotel_preference=hotel. The answer should address the booking or reservation step while tracking area, price, dates, people, and amenities. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target12 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"premature_respond"
] | 0.95 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
And when will you come back? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target20 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"fragment_answer",
"weak_final_state"
] | 0.499 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Could I interest you in our store credit card? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Can you tell me where the pots and pans are?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Pots and pans are right over there.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Oh, thank you.",
"turn_id": 3
}
] | DailyDialog_test_000102 | shopping | life_mt_sct_DailyDialog_test_000102_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"shopping_retail"
],
"domain_category"... | [
"financial_context",
"sensitive_content"
] | 0.749 | DailyDialog | test_000102 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
And what night will you be staying at the hotel, please? | The visible history is a hotel booking task in travel transportation. Known constraints are hotel_preference=hotel. The answer should address the booking or reservation step while tracking area, price, dates, people, and amenities. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target14 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [
"fragment_answer"
] | 0.499 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay. Do you want nonstop? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target22 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"weak_final_state"
] | 0.899 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
But our credit card saves you 10 percent. | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Can you tell me where the pots and pans are?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Pots and pans are right over there.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Oh, thank you.",
"turn_id": 3
},
{
"speaker": "assis... | DailyDialog_test_000102 | shopping | life_mt_sct_DailyDialog_test_000102_target6 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"shopping_retail"
],
"domain_category"... | [
"financial_context",
"fragment_answer",
"sensitive_content"
] | 0.499 | DailyDialog | test_000102 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Of course. What day will you be travelling? | The visible history is a train booking task in travel transportation. Known constraints are destination=broxbourne; hotel_preference=hotel; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target16 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 0.899 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay, there is actually one nonstop flight. It's with air France. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target24 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"premature_respond",
"weak_final_state"
] | 0.92 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Here. Let me give you an application form. | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Can you tell me where the pots and pans are?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Pots and pans are right over there.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Oh, thank you.",
"turn_id": 3
},
{
"speaker": "assis... | DailyDialog_test_000102 | shopping | life_mt_sct_DailyDialog_test_000102_target8 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"shopping_retail"
],
"domain_category"... | [
"financial_context",
"sensitive_content"
] | 0.749 | DailyDialog | test_000102 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
We have 14 trains leaving around 09:30 on Monday to Broxbourne. Was there a certain time you wanted to arrive by? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=monday, morning; destination=broxbourne; hotel_preference=hotel; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target18 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 1 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
It's $12091. Is that okay? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target26 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"weak_final_state"
] | 0.899 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay. Enjoy your browsing. | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Can you tell me where the pots and pans are?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Pots and pans are right over there.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Oh, thank you.",
"turn_id": 3
},
{
"speaker": "assis... | DailyDialog_test_000102 | shopping | life_mt_sct_DailyDialog_test_000102_target10 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"shopping_retail"
],
"domain_category"... | [
"financial_context",
"low_information_answer",
"sensitive_content"
] | 0.749 | DailyDialog | test_000102 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
The TR1319 fits your criteria. Would you like me to book passage for you on this train? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=monday, morning, 3 pm; destination=broxbourne; hotel_preference=hotel; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target20 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 1 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay. So, you will leave Los Angeles at 3:30 p.m. And arrive in Paris at 11:15 A.M. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target28 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"premature_respond",
"weak_final_state"
] | 0.92 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Booking was successful, the total fee is 71.59 GBP payable at the station. Reference number is: UYNXZ4W5. Do you need anything else? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=monday, morning, 3 pm; destination=broxbourne; hotel_preference=hotel; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target22 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 1 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
3:30 P.M. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target30 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"low_information_answer",
"weak_final_state"
] | 0.749 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay, great! I'm so glad I could help you. Have a wonderful visit! | The visible history is a train booking task in travel transportation. Known constraints are budget=4, 9; date_or_duration=monday, morning, 3 pm; destination=broxbourne; hotel_preference=hotel. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "I need a room in cambridge please",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Okay, do you have an area or price range you're looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "expensive guesthouse. it should have free parking ... | MultiWOZ_MUL2064.json | travel | life_mt_sct_MultiWOZ_MUL2064.json_target24 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"travel_tr... | [] | 1 | MultiWOZ | MUL2064.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
On your return you will leave at 1:10 P.M in Paris. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target32 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"premature_respond",
"weak_final_state"
] | 0.92 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Yes, how can I help you? | The dialogue is school-life in education career, with information request intent. Key constraints are deadline=morning. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "Good morning, may I speak with Professor Clark, please?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "You are speaking with Professor Clark.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Professor, I am Kalina from your morning literatur... | DailyDialog_test_000104 | study | life_mt_sct_DailyDialog_test_000104_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"education_career"
],
"domain_category... | [] | 0.899 | DailyDialog | test_000104 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Indeed, they do. City Centre North B&B is your best bet. Would you like to book a room there? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target4 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
And arrive in Los Angeles at 3:55 P.M. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target34 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"fragment_answer",
"premature_respond",
"weak_final_state"
] | 0.499 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Oh, my God! I hope you are all right. | The dialogue is school-life in education career, with planning coordination intent. Key constraints are deadline=morning. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "Good morning, may I speak with Professor Clark, please?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "You are speaking with Professor Clark.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Professor, I am Kalina from your morning literatur... | DailyDialog_test_000104 | study | life_mt_sct_DailyDialog_test_000104_target6 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"education_career"
],
"domain_category... | [
"answer_not_grounded",
"premature_respond"
] | 0.499 | DailyDialog | test_000104 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Free parking is included at the City Centre North B and B. Should I make a reservation for you? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target6 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Does this all look good? | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target36 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"weak_final_state"
] | 0.899 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
How much school will you miss? | The dialogue is school-life in education career, with information request intent. Key constraints are deadline=morning. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "Good morning, may I speak with Professor Clark, please?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "You are speaking with Professor Clark.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Professor, I am Kalina from your morning literatur... | DailyDialog_test_000104 | study | life_mt_sct_DailyDialog_test_000104_target8 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"education_career"
],
"domain_category... | [] | 0.899 | DailyDialog | test_000104 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Their address is 328a Histon Road. Anything else I can do for you? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target8 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Great! Let me send you the flight details. Bye. | The visible history is a movie ticketing task in travel transportation. It should track movie, showtime, and seats. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "I would like you to help me finding a flight.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Hi, how can I help you? Sorry, what did you say?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I said I would like you to help me in finding a fl... | Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | cooking | life_mt_sct_Taskmaster_dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6_target38 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"duplicate_turn",
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "... | [
"premature_respond",
"weak_final_state"
] | 0.899 | Taskmaster | dlg-02d7d49e-d5b7-4c8a-b982-e936506ddcf6 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: ingredients; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
I appreciate you calling and telling me that you won ' t be in class. See you next week! | The dialogue is school-life in education career, with planning coordination intent. Key constraints are deadline=morning. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "Good morning, may I speak with Professor Clark, please?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "You are speaking with Professor Clark.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Professor, I am Kalina from your morning literatur... | DailyDialog_test_000104 | study | life_mt_sct_DailyDialog_test_000104_target10 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"education_career"
],
"domain_category... | [
"premature_respond"
] | 0.95 | DailyDialog | test_000104 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: deadline; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Do you like thai food? The sala thong is a good restaurant that meets your needs. Phone is: [phone removed] Postcode: cb39ey | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target10 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Let me see. I can come out and see you first thing Wednesday. | The dialogue is schedule planning in personal schedule, with planning coordination intent. Key constraints are date_or_time=next week. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "Hello, Parker. How ' s everything?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Can ' t complain. And you?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Business is booming. I understand you want to meet up with me next week. How ' s yo... | DailyDialog_test_000105 | schedule | life_mt_sct_DailyDialog_test_000105_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"personal_... | [
"answer_not_grounded",
"off_topic_answer",
"premature_respond"
] | 0.25 | DailyDialog | test_000105 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Where will you be headed and when would you like to depart by? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target12 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
I have the same question. I can ' t send out mails. We ' d better call the IT department and ask them to check it immediately. | The dialogue is workplace conversation in education career, with negotiation decision intent. The most recent user context is: Yes, I wanted to look for some information on the company page just now. It took me almost one minute to open it. Then there is no response to any click. The answer is grounded only in the visi... | [
{
"speaker": "user",
"text": "How come it is slow as a snail today?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "You mean the network connection?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Yes, I wanted to look for some information on the company page just now. I... | DailyDialog_test_000106 | other_life | life_mt_sct_DailyDialog_test_000106_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"education... | [
"answer_not_grounded",
"off_topic_answer",
"weak_final_state"
] | 0.27 | DailyDialog | test_000106 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Great Your Booking is complete on a yellow lexus. The Contact number is [phone removed]. Anything else I can help you with? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target14 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Dad, I don't want to talk about it now. | The dialogue is school-life in social relationship, with information request intent. The most recent user context is: No, honey, why didn't you go to cram school last night? The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "Honey, I need to have a talk with you.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Dad, I have to do my homework.",
"turn_id": 2
},
{
"speaker": "user",
"text": "No, honey, why didn't you go to cram school last night?",
"turn_id": 3
... | DailyDialog_test_000107 | study | life_mt_sct_DailyDialog_test_000107_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"premature_respond",
"weak_final_state"
] | 0.899 | DailyDialog | test_000107 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Thanks for using the Cambridge Town Info centre Services! | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "Yes, hello. I need a place to crash so I'm thinking a 0-star guesthouse",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "great, i have two options for you",
"turn_id": 2
},
{
"speaker": "user",
"text": "Ill take either one but can you check a... | MultiWOZ_PMUL4581.json | travel | life_mt_sct_MultiWOZ_PMUL4581.json_target16 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"financial_context",
"premature_respond",
"sensitive_content"
] | 0.749 | MultiWOZ | PMUL4581.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
I'm sorry, dad. But I would rather stay at school than go to cram school. | The dialogue is school-life in social relationship, with booking or reservation intent. The most recent user context is: Honey, if you don't want to go to cram school, you should tell me the reason why. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "Honey, I need to have a talk with you.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Dad, I have to do my homework.",
"turn_id": 2
},
{
"speaker": "user",
"text": "No, honey, why didn't you go to cram school last night?",
"turn_id": 3
... | DailyDialog_test_000107 | study | life_mt_sct_DailyDialog_test_000107_target6 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"premature_respond",
"weak_final_state"
] | 0.92 | DailyDialog | test_000107 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
I don't know, but I feel terrible. | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "What's wrong with you, young man?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Doctor, I have a bad cough and a headache.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Do you have a fever?",
"turn_id": 3
}
] | DailyDialog_test_000108 | health_routine | life_mt_sct_DailyDialog_test_000108_target4 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"health_wellness"
],
"domain_category"... | [
"medical_or_health_context",
"sensitive_content"
] | 0.749 | DailyDialog | test_000108 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Do you think I should lie in bed? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "What's wrong with you, young man?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Doctor, I have a bad cough and a headache.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Do you have a fever?",
"turn_id": 3
},
{
"speaker": "ass... | DailyDialog_test_000108 | health_routine | life_mt_sct_DailyDialog_test_000108_target6 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"health_wellness"
],
"domain_category"... | [
"medical_or_health_context",
"sensitive_content"
] | 0.749 | DailyDialog | test_000108 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Do you want to leave in the afternoon or evening flight? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=wednesday; origin=sacramento; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Hi. I need a flight today from Sacramento to Denver.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What are those dates?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Today, and then I'll be back on Wednesday.",
"turn_id": 3
}
] | Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | travel | life_mt_sct_Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7_target4 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
OK. Do you think I can play football tomorrow? | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "What's wrong with you, young man?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Doctor, I have a bad cough and a headache.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Do you have a fever?",
"turn_id": 3
},
{
"speaker": "ass... | DailyDialog_test_000108 | health_routine | life_mt_sct_DailyDialog_test_000108_target8 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"personal_schedule"
],
"domain_categor... | [
"medical_or_health_context",
"sensitive_content"
] | 0.749 | DailyDialog | test_000108 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Ok. What seating class do you prefer? What seating class do you prefer? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=wednesday, evening; origin=sacramento; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Hi. I need a flight today from Sacramento to Denver.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What are those dates?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Today, and then I'll be back on Wednesday.",
"turn_id": 3
},
{... | Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | travel | life_mt_sct_Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7_target6 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
OK, I'll listen to you. | This sample contains safety-sensitive context. The reasoning should stay conservative, avoid amplifying sensitive details, and only support a safe, bounded assistant response. | [
{
"speaker": "user",
"text": "What's wrong with you, young man?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Doctor, I have a bad cough and a headache.",
"turn_id": 2
},
{
"speaker": "user",
"text": "Do you have a fever?",
"turn_id": 3
},
{
"speaker": "ass... | DailyDialog_test_000108 | health_routine | life_mt_sct_DailyDialog_test_000108_target10 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"personal_schedule"
],
"domain_categor... | [
"low_information_answer",
"medical_or_health_context",
"sensitive_content"
] | 0.749 | DailyDialog | test_000108 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the visible dialogue prefix is used. The turn is handled conservatively because the context is safety-sensitive, so the state keeps prior safe context and waits for a safe next response.",
"reasoning_confidence": "medium",
"reasoning_type": "safet... | prefix_to_next_assistant_text |
Do you prefer non stop or layover? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=wednesday, evening; origin=sacramento; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Hi. I need a flight today from Sacramento to Denver.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What are those dates?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Today, and then I'll be back on Wednesday.",
"turn_id": 3
},
{... | Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | travel | life_mt_sct_Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7_target8 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 0.899 | Taskmaster | dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
What kind are you thinking of getting? | The dialogue is restaurant search in food dining, with information request intent. Key constraints are meal_type=dinner, dessert. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "I want something sweet after dinner.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What do you have in mind?",
"turn_id": 2
},
{
"speaker": "user",
"text": "A dessert sounds nice.",
"turn_id": 3
}
] | DailyDialog_test_000109 | cooking | life_mt_sct_DailyDialog_test_000109_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dining"
],... | [] | 0.899 | DailyDialog | test_000109 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | prefix_to_next_assistant_text |
Ok. I found a flight at United. departs at 5:55 P.M. – arrives at 9:18 P.M. your return flight at departs at 7:21 P.M. – arrives at 8:55 P.M. I found a flight at United. departs at 5:55 PM – arrives at 9:18 PM. your return flight at departs at 7:21 PM – arrives at 8:55 PM. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=wednesday, evening; origin=sacramento; transport_type=flight. The answer should address the customer support step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Hi. I need a flight today from Sacramento to Denver.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What are those dates?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Today, and then I'll be back on Wednesday.",
"turn_id": 3
},
{... | Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | travel | life_mt_sct_Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7_target10 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [
"premature_respond"
] | 0.95 | Taskmaster | dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
What kind of pie do you want? | The dialogue is restaurant search in food dining, with information request intent. Key constraints are meal_type=dinner, dessert. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "I want something sweet after dinner.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What do you have in mind?",
"turn_id": 2
},
{
"speaker": "user",
"text": "A dessert sounds nice.",
"turn_id": 3
},
{
"speaker": "assistant",
... | DailyDialog_test_000109 | cooking | life_mt_sct_DailyDialog_test_000109_target6 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dining"
],... | [] | 0.899 | DailyDialog | test_000109 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | prefix_to_next_assistant_text |
Certainly. They're at 4 - 6 Rose Crescent. Their number is [phone removed]. I can help you book a table if you'd like. Or is there anything else you need? | The visible history is a restaurant booking task in travel transportation. It should track cuisine, area, price, time, and party size. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "Do you have information on La Raza?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "La raza is a spanish restaurant in the centre area. It is considered cheap",
"turn_id": 2
},
{
"speaker": "user",
"text": "Can I get their address and phone ... | MultiWOZ_MUL1603.json | travel | life_mt_sct_MultiWOZ_MUL1603.json_target4 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"weak_final_state"
] | 0.97 | MultiWOZ | MUL1603.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Ok. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=wednesday, evening; origin=sacramento; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Hi. I need a flight today from Sacramento to Denver.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What are those dates?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Today, and then I'll be back on Wednesday.",
"turn_id": 3
},
{... | Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | travel | life_mt_sct_Taskmaster_dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7_target12 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [
"answer_not_grounded",
"fragment_answer",
"generic_answer",
"too_short_answer",
"weak_target_answer"
] | 0 | Taskmaster | dlg-032d1836-0f6e-4826-bdd2-51ae123dcdd7 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: origin, transport_type; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_con... | prefix_to_next_assistant_text |
Do you want to know what kind of pie I like? | The dialogue is restaurant search in food dining, with confirmation clarification intent. Key constraints are meal_type=dinner, dessert. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "I want something sweet after dinner.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What do you have in mind?",
"turn_id": 2
},
{
"speaker": "user",
"text": "A dessert sounds nice.",
"turn_id": 3
},
{
"speaker": "assistant",
... | DailyDialog_test_000109 | cooking | life_mt_sct_DailyDialog_test_000109_target8 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dining"
],... | [] | 1 | DailyDialog | test_000109 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | prefix_to_next_assistant_text |
There are 278 trains that leave after 11:00 on Saturday. Do you have a particular departure and arrival location? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=saturday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Do you have information on La Raza?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "La raza is a spanish restaurant in the centre area. It is considered cheap",
"turn_id": 2
},
{
"speaker": "user",
"text": "Can I get their address and phone ... | MultiWOZ_MUL1603.json | travel | life_mt_sct_MultiWOZ_MUL1603.json_target6 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [] | 1 | MultiWOZ | MUL1603.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Do you have a preference, which New York airport? | The visible history is a flight booking task in travel transportation. Known constraints are destination=seattle; transport_type=flight. The answer should address the confirmation clarification step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target4 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 0.899 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
I love apple pie. | The dialogue is restaurant search in food dining, with information request intent. Key constraints are meal_type=dinner, dessert. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "I want something sweet after dinner.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What do you have in mind?",
"turn_id": 2
},
{
"speaker": "user",
"text": "A dessert sounds nice.",
"turn_id": 3
},
{
"speaker": "assistant",
... | DailyDialog_test_000109 | cooking | life_mt_sct_DailyDialog_test_000109_target10 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dining"
],... | [
"low_information_answer"
] | 0.749 | DailyDialog | test_000109 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | prefix_to_next_assistant_text |
It looks like train TR6210 will be leaving Leicester at 11:09 on Saturday and will arrive to Cambridge at 12:54. Can I help you with anything else? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=saturday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Do you have information on La Raza?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "La raza is a spanish restaurant in the centre area. It is considered cheap",
"turn_id": 2
},
{
"speaker": "user",
"text": "Can I get their address and phone ... | MultiWOZ_MUL1603.json | travel | life_mt_sct_MultiWOZ_MUL1603.json_target8 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [] | 1 | MultiWOZ | MUL1603.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay, got it. When would you like to depart and return? | The visible history is a flight booking task in travel transportation. Known constraints are destination=seattle; transport_type=flight. The answer should address the customer support step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target6 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
There you go. Problem solved. | The dialogue is restaurant search in food dining, with problem solving intent. Key constraints are meal_type=dinner, dessert. The answer is grounded because it responds to that goal using only those stated details. | [
{
"speaker": "user",
"text": "I want something sweet after dinner.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "What do you have in mind?",
"turn_id": 2
},
{
"speaker": "user",
"text": "A dessert sounds nice.",
"turn_id": 3
},
{
"speaker": "assistant",
... | DailyDialog_test_000109 | cooking | life_mt_sct_DailyDialog_test_000109_target12 | false | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dining"
],... | [
"low_information_answer"
] | 0.749 | DailyDialog | test_000109 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: meal_type; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",... | prefix_to_next_assistant_text |
The travel time for that route is 105 minutes. Is there anything else I can help you with? | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=saturday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Do you have information on La Raza?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "La raza is a spanish restaurant in the centre area. It is considered cheap",
"turn_id": 2
},
{
"speaker": "user",
"text": "Can I get their address and phone ... | MultiWOZ_MUL1603.json | travel | life_mt_sct_MultiWOZ_MUL1603.json_target10 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [] | 1 | MultiWOZ | MUL1603.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Okay, so you like to leave Sunday May 7th in the evening and return Saturday May 13th. Do you have a preference for time on Saturday to return? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the customer support step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target8 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
Glad I could be of help. Have a great day. Goodbye. | The visible history is a train booking task in travel transportation. Known constraints are date_or_duration=saturday; transport_type=train. The answer should address the booking or reservation step while tracking departure, destination, and time constraints. | [
{
"speaker": "user",
"text": "Do you have information on La Raza?",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "La raza is a spanish restaurant in the centre area. It is considered cheap",
"turn_id": 2
},
{
"speaker": "user",
"text": "Can I get their address and phone ... | MultiWOZ_MUL1603.json | travel | life_mt_sct_MultiWOZ_MUL1603.json_target12 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"premature_respond"
] | 0.95 | MultiWOZ | MUL1603.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: destination; resolved missing slots: none.",
"reasoning_confidence": "high",
... | prefix_to_next_assistant_text |
Got it. How many tickets do you need? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target10 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 0.899 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
Do you have other preferences for the flight? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target12 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 0.899 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
Okay, anything more? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target14 | false | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [
"too_short_answer",
"weak_target_answer"
] | 0.68 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
Okay, I only have one flight available with American. Will that airline be suitable? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target16 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
Sure. American Airlines flight leaves at 5:29 p.m. on May 7th and arrives in Seattle at 9 p.m., it is a nonstop flight first class with Wi-Fi. Does this suit your needs? | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the booking or reservation step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target18 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
There are 4 museums in the east: Cambridge Artworks, Cambridge Museum of Technology, Gallery at Twelve a High Street, and Saint Barnabas Press Gallery. | The visible history is a attraction search task in travel transportation. It should track the visible goal, constraints, and next step. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "Hi, I'm looking for a place to go in the east.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "ok, what type of place are you looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Hi im looking for a museum",
"turn_id": 3
}
] | MultiWOZ_MUL2393.json | schedule | life_mt_sct_MultiWOZ_MUL2393.json_target4 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"premature_respond",
"weak_final_state"
] | 0.92 | MultiWOZ | MUL2393.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: date_or_time, event_type; resolved missing slots: none.",
"reasoning_confidence... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: date_or_time, event_type; resolved missing slots: none.",
"reasoning_confidence... | prefix_to_next_assistant_text |
Okay, let me get your return flight details. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the confirmation clarification step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target20 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 0.899 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
Of course. The Phone number is [phone removed] and postode is cb13ef. What else can I help you with? | The visible history is a attraction search task in travel transportation. It should track the visible goal, constraints, and next step. The answer should ask, confirm, or respond only to details present in the visible history. | [
{
"speaker": "user",
"text": "Hi, I'm looking for a place to go in the east.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "ok, what type of place are you looking for?",
"turn_id": 2
},
{
"speaker": "user",
"text": "Hi im looking for a museum",
"turn_id": 3
},
{... | MultiWOZ_MUL2393.json | schedule | life_mt_sct_MultiWOZ_MUL2393.json_target6 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"food_dini... | [
"weak_final_state"
] | 0.97 | MultiWOZ | MUL2393.json | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: date_or_time, event_type; resolved missing slots: none.",
"reasoning_confidence... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible multiwoz prefix is used. The user turn is treated as new task evidence; added slots: none; changed slots: none; removed slots: none; new missing slots: date_or_time, event_type; resolved missing slots: none.",
"reasoning_confidence... | prefix_to_next_assistant_text |
One moment. Return flight attached to this is a separate ticket on Alaska Airlines but it is also non-stop first class with Wi-Fi. It leaves at 9:30 p.m. and arrives in Seattle at 6 a.m. | The visible history is a flight booking task in travel transportation. Known constraints are date_or_duration=sunday, evening, 1 week; destination=seattle; transport_type=flight. The answer should address the customer support step while tracking origin, destination, and travel dates. | [
{
"speaker": "user",
"text": "Yes. I'm looking for a flight. I'm going to Seattle.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "How can I help you? Great. I can help you book a flight. I see that you're in New York, is that where you want to fly out from?",
"turn_id": 2
},
{
... | Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | travel | life_mt_sct_Taskmaster_dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d_target22 | true | {
"category": "task_oriented_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [
"too_many_turns",
"too_short_average_turn"
],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain... | [] | 1 | Taskmaster | dlg-0331ef03-09fb-4272-9b2e-c1ef16f4be4d | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible taskmaster prefix is used. The user turn is treated as new task evidence; added slots: destination, transport_type; changed slots: none; removed slots: none; new missing slots: none; resolved missing slots: none.",
"reasoning_confi... | prefix_to_next_assistant_text |
I see. Let me show you some samples of various qualities. This one is very nice. It's three hundred and fifty dollars. | The dialogue is family conversation in social relationship, with information request intent. The most recent user context is: I'm not sure. I don't know very much about the price of pearls. The answer is grounded only in the visible dialogue history and should stay conservative. | [
{
"speaker": "user",
"text": "I'm looking for a string of pearls for my wife.",
"turn_id": 1
},
{
"speaker": "assistant",
"text": "Yes, sir. What price range do you have in mind?",
"turn_id": 2
},
{
"speaker": "user",
"text": "I'm not sure. I don't know very much about the pr... | DailyDialog_test_000113 | shopping | life_mt_sct_DailyDialog_test_000113_target4 | true | {
"category": "daily_dialogue",
"category_taxonomy_method": "deterministic_source_aware_rules_v1",
"chunk_level": "turn",
"cleaning_flags": [],
"construction_type": "prefix_to_next_assistant_turn",
"deep_reasoning_method": "rule_based",
"domain_categories": [
"social_relationship"
],
"domain_categ... | [
"premature_respond",
"weak_final_state"
] | 0.92 | DailyDialog | test_000113 | train | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | [
{
"chunk_id": "C1",
"reasoning": "After turn 1, only the current visible dailydialog prefix is used. The user turn updates the daily dialogue context and response policy without relying on future turns; added slots: none; changed slots: none; removed slots: none.",
"reasoning_confidence": "medium",
... | prefix_to_next_assistant_text |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.