id
stringlengths 1
12
| question
stringlengths 22
566
| golden_answers
dict | data_source
stringclasses 2
values | prompt
listlengths 1
1
| ability
stringclasses 6
values | reward_model
dict | extra_info
dict | metadata
dict |
|---|---|---|---|---|---|---|---|---|
exfever_8968
|
Céline Marie Claudette Dion is a Canadian singer born into a large family from an off-island suburb of Montreal, Quebec, Canada.
|
{
"evidence": [
"Celine Dion",
"Charlemagne, Quebec"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCéline Marie Claudette Dion is a Canadian singer born into a large family from an off-island suburb of Montreal, Quebec, Canada.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Celine Dion",
"Charlemagne, Quebec"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7484,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6336
|
Destiny is an online-only multiplayer game that is centered on guns and other weapon-based combat from a first-person perspective.
|
{
"evidence": [
"Destiny (video game)",
"First-person shooter"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nDestiny is an online-only multiplayer game that is centered on guns and other weapon-based combat from a first-person perspective.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Destiny (video game)",
"First-person shooter"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7485,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5118
|
Adrienne Bailon is an American television host and a former member of an American boy group consisting of talented singers, including an American television host, singer, and actress.
|
{
"evidence": [
"Adrienne Bailon",
"The Cheetah Girls (group)",
"Adrienne Bailon"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAdrienne Bailon is an American television host and a former member of an American boy group consisting of talented singers, including an American television host, singer, and actress.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Adrienne Bailon",
"The Cheetah Girls (group)",
"Adrienne Bailon"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7486,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5144
|
One Direction are an English-Irish pop boy band formed in London, England in 2011, which signed with the record label Syco Records by an English television personality who also judged The X Factor UK.
|
{
"evidence": [
"One Direction",
"Simon Cowell"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nOne Direction are an English-Irish pop boy band formed in London, England in 2011, which signed with the record label Syco Records by an English television personality who also judged The X Factor UK.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"One Direction",
"Simon Cowell"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7487,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8834
|
The Canadian dollar has been issued by chartered banks since the 1830s.
|
{
"evidence": [
"Omar Khadr",
"Canadian dollar"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Canadian dollar has been issued by chartered banks since the 1830s.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Omar Khadr",
"Canadian dollar"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7488,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
14443
|
Akins attended Cleveland High School in Los Angeles, California, then he played for the United States national baseball team in the 1984 Winter Olympics.
|
{
"evidence": [
"Sid Akins_sentence_2"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAkins attended Cleveland High School in Los Angeles, California, then he played for the United States national baseball team in the 1984 Winter Olympics.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Sid Akins_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7489,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.586",
"value": "?search="
},
{
"operation": "search",
"time": "24.075",
"value": "United States national baseball team in the 1984 Winter Olympics"
},
{
"operation": "Now on",
"time": "27.223",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "72.023",
"value": "Nations at the Winter Olympics Olympic Games_section_42"
},
{
"operation": "Now on",
"time": "72.745",
"value": "Olympic Games#Nations at the Winter Olympics Olympic Games section 42"
},
{
"operation": "search",
"time": "103.304",
"value": "1984 Winter Olympics"
},
{
"operation": "Now on",
"time": "104.189",
"value": "1984 Winter Olympics"
},
{
"operation": "search",
"time": "161.351",
"value": "Akins United States national baseball team"
},
{
"operation": "Now on",
"time": "162.384",
"value": "undefined"
},
{
"operation": "search",
"time": "180.934",
"value": "1984 United States national baseball team"
},
{
"operation": "Now on",
"time": "182.049",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "185.092",
"value": "United States national baseball team"
},
{
"operation": "Now on",
"time": "186.013",
"value": "United States national baseball team"
},
{
"operation": "hyperlink",
"time": "241.147",
"value": "1984"
},
{
"operation": "Now on",
"time": "242.335",
"value": "Baseball at the 1984 Summer Olympics"
},
{
"operation": "hyperlink",
"time": "260.699",
"value": "United States"
},
{
"operation": "Now on",
"time": "261.113",
"value": "United States national baseball team"
},
{
"operation": "back-button-clicked",
"time": "290.508",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "290.683",
"value": "Baseball at the 1984 Summer Olympics"
},
{
"operation": "hyperlink",
"time": "305.45",
"value": "United States"
},
{
"operation": "Now on",
"time": "305.835",
"value": "United States national baseball team"
},
{
"operation": "search",
"time": "399.567",
"value": "sid akins"
},
{
"operation": "Now on",
"time": "400.488",
"value": "Sid Akins"
},
{
"operation": "Highlighting",
"time": "404.321",
"value": "Sid Akins_sentence_2"
},
{
"operation": "finish",
"time": "426.727",
"value": "finish"
}
]
}
|
62428
|
On 30 January 2008, Bulgarian National Television presented its answer to the BBC documentary in the form of a documentary called Dom (House or Home); as a result, on the same day, Bulgarian Members of the European Parliament signed a declaration in which they called for "a common European policy on the upbringing and education of underprivileged children."
|
{
"evidence": [
"Mogilino_sentence_7",
"Mogilino_sentence_9"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nOn 30 January 2008, Bulgarian National Television presented its answer to the BBC documentary in the form of a documentary called Dom (House or Home); as a result, on the same day, Bulgarian Members of the European Parliament signed a declaration in which they called for \"a common European policy on the upbringing and education of underprivileged children.\"\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Mogilino_sentence_7",
"Mogilino_sentence_9"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7490,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.127",
"value": "?search="
},
{
"operation": "search",
"time": "97.824",
"value": "Bulgarian National Television"
},
{
"operation": "Now on",
"time": "98.569",
"value": "Bulgarian National Television"
},
{
"operation": "search",
"time": "149.742",
"value": "Dom"
},
{
"operation": "Now on",
"time": "150.617",
"value": "Dom"
},
{
"operation": "search",
"time": "191.682",
"value": "dom House or Home"
},
{
"operation": "Now on",
"time": "194.078",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "196.943",
"value": "Mogilino"
},
{
"operation": "Now on",
"time": "197.378",
"value": "Mogilino"
},
{
"operation": "Highlighting",
"time": "202.799",
"value": "Mogilino_sentence_7"
},
{
"operation": "Highlighting",
"time": "303.053",
"value": "Mogilino_sentence_9"
},
{
"operation": "finish",
"time": "339.519",
"value": "finish"
}
]
}
|
60063
|
Robert Scott Moncrieff, who died in Edinburgh on 18 June 1869, is buried with other members of the Scott Moncrieff family at the south end of the sealed south-west section of Greyfriars Kirkyard commonly called the Covenanter's Prison.
|
{
"evidence": [
"Robert Scott Moncrieff_sentence_6",
"Robert Scott Moncrieff_sentence_7"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nRobert Scott Moncrieff, who died in Edinburgh on 18 June 1869, is buried with other members of the Scott Moncrieff family at the south end of the sealed south-west section of Greyfriars Kirkyard commonly called the Covenanter's Prison.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Robert Scott Moncrieff_sentence_6",
"Robert Scott Moncrieff_sentence_7"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7491,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.464",
"value": "?search="
},
{
"operation": "search",
"time": "33.531",
"value": "Robert Scott Moncrieff"
},
{
"operation": "Now on",
"time": "34.288",
"value": "Robert Scott Moncrieff"
},
{
"operation": "Highlighting",
"time": "42.258",
"value": "Robert Scott Moncrieff_sentence_6"
},
{
"operation": "Highlighting",
"time": "46.908",
"value": "Robert Scott Moncrieff_sentence_7"
},
{
"operation": "finish",
"time": "277.732",
"value": "finish"
}
]
}
|
58126
|
The Alessandria–Piacenza railway first connected Alessandria, first founded in 1168, and Casteggio and eventually connected to Rome.
|
{
"evidence": [
"Alessandria–Piacenza railway_sentence_6",
"Alessandria_sentence_4"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Alessandria–Piacenza railway first connected Alessandria, first founded in 1168, and Casteggio and eventually connected to Rome.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Alessandria–Piacenza railway_sentence_6",
"Alessandria_sentence_4"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7492,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.651",
"value": "?search="
},
{
"operation": "search",
"time": "11.199",
"value": "Alessandria–Piacenza railway"
},
{
"operation": "Now on",
"time": "12.408",
"value": "Alessandria–Piacenza railway"
},
{
"operation": "Page search",
"time": "38.986",
"value": "Alessandria"
},
{
"operation": "search",
"time": "50.906",
"value": "Alessandria"
},
{
"operation": "Now on",
"time": "53.015",
"value": "Alessandria"
},
{
"operation": "back-button-clicked",
"time": "65.429",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "65.725",
"value": "Alessandria–Piacenza railway"
},
{
"operation": "Highlighting",
"time": "93.273",
"value": "Alessandria–Piacenza railway_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "115.51",
"value": "Alessandria–Piacenza railway_sentence_3"
},
{
"operation": "Highlighting",
"time": "122.245",
"value": "Alessandria–Piacenza railway_sentence_6"
},
{
"operation": "hyperlink",
"time": "132.398",
"value": "Casteggio"
},
{
"operation": "Now on",
"time": "133.17",
"value": "Casteggio"
},
{
"operation": "Page search",
"time": "167.98",
"value": "founded"
},
{
"operation": "back-button-clicked",
"time": "172.528",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "172.787",
"value": "Alessandria–Piacenza railway"
},
{
"operation": "Highlighting deleted",
"time": "175.672",
"value": "Alessandria–Piacenza railway_sentence_6"
},
{
"operation": "Highlighting",
"time": "176.567",
"value": "Alessandria–Piacenza railway_sentence_6"
},
{
"operation": "Highlighting deleted",
"time": "184.463",
"value": "Alessandria–Piacenza railway_sentence_6"
},
{
"operation": "Highlighting",
"time": "186.439",
"value": "Alessandria–Piacenza railway_sentence_6"
},
{
"operation": "search",
"time": "205.094",
"value": "Alessandria"
},
{
"operation": "Now on",
"time": "206.905",
"value": "Alessandria"
},
{
"operation": "Highlighting",
"time": "212.505",
"value": "Alessandria_sentence_4"
},
{
"operation": "back-button-clicked",
"time": "215.001",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "215.266",
"value": "Alessandria–Piacenza railway"
},
{
"operation": "Page search",
"time": "228.103",
"value": "Rome"
},
{
"operation": "Highlighting",
"time": "251.794",
"value": "Alessandria–Piacenza railway_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "258.106",
"value": "Alessandria–Piacenza railway_sentence_3"
},
{
"operation": "finish",
"time": "315.339",
"value": "finish"
}
]
}
|
53723
|
Hassan Illiassou competed in the 100 m event, but only ranked fifth with a result of 10.73, while Boureima Kimba was the only one to advance to the quarterfinals.
|
{
"evidence": [
"Hassane Illiassou_sentence_1",
"Hassane Illiassou_sentence_0",
"Bouriema Kimba_sentence_0"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHassan Illiassou competed in the 100 m event, but only ranked fifth with a result of 10.73, while Boureima Kimba was the only one to advance to the quarterfinals.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Hassane Illiassou_sentence_1",
"Hassane Illiassou_sentence_0",
"Bouriema Kimba_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7493,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.75",
"value": "?search="
},
{
"operation": "search",
"time": "11.944",
"value": "Hassan Illiassou"
},
{
"operation": "Now on",
"time": "12.754",
"value": "Hassane Illiassou"
},
{
"operation": "Highlighting",
"time": "19.572",
"value": "Hassane Illiassou_sentence_0"
},
{
"operation": "Highlighting",
"time": "33.767",
"value": "Hassane Illiassou_sentence_1"
},
{
"operation": "hyperlink",
"time": "46.005",
"value": "1992 Summer Olympics"
},
{
"operation": "Now on",
"time": "73.851",
"value": "1992 Summer Olympics"
},
{
"operation": "back-button-clicked",
"time": "79.001",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "79.441",
"value": "Hassane Illiassou"
},
{
"operation": "Highlighting deleted",
"time": "92.99",
"value": "Hassane Illiassou_sentence_0"
},
{
"operation": "Highlighting",
"time": "97.795",
"value": "Hassane Illiassou_sentence_0"
},
{
"operation": "search",
"time": "102.654",
"value": "Contains... men's 100m competition at the 1992 Summer Olympics"
},
{
"operation": "Now on",
"time": "106.606",
"value": "undefined"
},
{
"operation": "search",
"time": "138.155",
"value": "Boureima Kimba"
},
{
"operation": "Now on",
"time": "138.836",
"value": "Bouriema Kimba"
},
{
"operation": "Highlighting",
"time": "145.986",
"value": "Bouriema Kimba_sentence_0"
},
{
"operation": "hyperlink",
"time": "150.968",
"value": "1996 Summer Olympics"
},
{
"operation": "Now on",
"time": "164.958",
"value": "1996 Summer Olympics"
},
{
"operation": "finish",
"time": "185.386",
"value": "finish"
}
]
}
|
exfever_7756
|
Randy Orton is an American professional wrestler and actor who is the son of an American semi-retired professional wrestler, better known by the ring name "Cowboy".
|
{
"evidence": [
"Randy Orton",
"Bob Orton Jr."
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nRandy Orton is an American professional wrestler and actor who is the son of an American semi-retired professional wrestler, better known by the ring name \"Cowboy\".\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Randy Orton",
"Bob Orton Jr."
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7494,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6597
|
Cosmetics are constituted mixtures of chemical compounds derived from a substance produced by a living organism, which can also be prepared by the artificial execution of chemical reactions to obtain several products.
|
{
"evidence": [
"Cosmetics",
"Natural product",
"Chemical synthesis"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCosmetics are constituted mixtures of chemical compounds derived from a substance produced by a living organism, which can also be prepared by the artificial execution of chemical reactions to obtain several products.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Cosmetics",
"Natural product",
"Chemical synthesis"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7495,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
92168
|
2044 Wirt is a minor planet named after its discoverer Carl Wirtanen in Novermber 8, 1950 at the Lick Observatory, an astronomical observatory owned and operated by Caltech atop the summit of Mount Hamilton,
|
{
"evidence": [
"Lick Observatory_sentence_0"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\n2044 Wirt is a minor planet named after its discoverer Carl Wirtanen in Novermber 8, 1950 at the Lick Observatory, an astronomical observatory owned and operated by Caltech atop the summit of Mount Hamilton,\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Lick Observatory_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7496,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "8.043",
"value": "undefined"
},
{
"operation": "search",
"time": "15.407",
"value": "2044 Wirt"
},
{
"operation": "Now on",
"time": "17.254",
"value": "2044 Wirt"
},
{
"operation": "Highlighting",
"time": "29.412",
"value": "2044 Wirt_cell_0_6_1"
},
{
"operation": "Highlighting",
"time": "29.414",
"value": "2044 Wirt_title"
},
{
"operation": "Highlighting",
"time": "29.448",
"value": "2044 Wirt_header_cell_0_6_0"
},
{
"operation": "Highlighting",
"time": "29.448",
"value": "2044 Wirt_header_cell_0_4_0"
},
{
"operation": "Highlighting",
"time": "33.538",
"value": "2044 Wirt_cell_0_3_1"
},
{
"operation": "Highlighting",
"time": "33.571",
"value": "2044 Wirt_header_cell_0_3_0"
},
{
"operation": "Highlighting",
"time": "33.571",
"value": "2044 Wirt_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "40.181",
"value": "2044 Wirt_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "42.197",
"value": "2044 Wirt_cell_0_3_1"
},
{
"operation": "Highlighting deleted",
"time": "42.226",
"value": "2044 Wirt_header_cell_0_3_0"
},
{
"operation": "Highlighting deleted",
"time": "42.227",
"value": "2044 Wirt_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "65.924",
"value": "2044 Wirt_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "71.398",
"value": "2044 Wirt_sentence_1"
},
{
"operation": "Highlighting",
"time": "76.604",
"value": "2044 Wirt_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "82.827",
"value": "2044 Wirt_cell_0_6_1"
},
{
"operation": "Highlighting deleted",
"time": "82.853",
"value": "2044 Wirt_header_cell_0_6_0"
},
{
"operation": "Highlighting deleted",
"time": "82.854",
"value": "2044 Wirt_header_cell_0_4_0"
},
{
"operation": "hyperlink",
"time": "87.13",
"value": "Lick Observatory"
},
{
"operation": "Now on",
"time": "88.859",
"value": "Lick Observatory"
},
{
"operation": "Highlighting",
"time": "95.022",
"value": "Lick Observatory_sentence_0"
},
{
"operation": "hyperlink",
"time": "102.249",
"value": "University of California"
},
{
"operation": "Now on",
"time": "104.716",
"value": "University of California"
},
{
"operation": "back-button-clicked",
"time": "109.96",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "110.109",
"value": "Lick Observatory"
},
{
"operation": "finish",
"time": "120.978",
"value": "finish"
}
]
}
|
exfever_4819
|
John Mayer attended a private music college in Boston, Massachusetts.
|
{
"evidence": [
"John Mayer",
"Berklee College of Music"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJohn Mayer attended a private music college in Boston, Massachusetts.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"John Mayer",
"Berklee College of Music"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7497,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5468
|
The Resident Evil film was bought by the German production company Constantin Film.
|
{
"evidence": [
"Resident Evil (film)",
"Colin Salmon",
"James Bond in film"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Resident Evil film was bought by the German production company Constantin Film.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Resident Evil (film)",
"Colin Salmon",
"James Bond in film"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7498,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
90850
|
India and China have a strained relationship especially with political and border issues which has made treaty adjustments difficult to finalize.
|
{
"evidence": [
"India-China border infrastructure_sentence_1",
"China–India relations_sentence_1"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nIndia and China have a strained relationship especially with political and border issues which has made treaty adjustments difficult to finalize.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"India-China border infrastructure_sentence_1",
"China–India relations_sentence_1"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7499,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.8",
"value": "?search="
},
{
"operation": "search",
"time": "35.45",
"value": "foreign relations with china"
},
{
"operation": "Now on",
"time": "37.618",
"value": "undefined"
},
{
"operation": "search",
"time": "50.139",
"value": "China-India Relations"
},
{
"operation": "Now on",
"time": "51.437",
"value": "China–India relations"
},
{
"operation": "Page search",
"time": "55.979",
"value": "strain"
},
{
"operation": "search",
"time": "87.144",
"value": "Foreign relations of China"
},
{
"operation": "Now on",
"time": "88.666",
"value": "Foreign relations of China"
},
{
"operation": "Page search",
"time": "91.597",
"value": "strained"
},
{
"operation": "Page search",
"time": "129.288",
"value": "india"
},
{
"operation": "hyperlink",
"time": "146.92",
"value": "7 Relations by region and country Foreign relations of China_section_9"
},
{
"operation": "search",
"time": "214.677",
"value": "India-China border"
},
{
"operation": "Now on",
"time": "215.8",
"value": "India-China border"
},
{
"operation": "Page search",
"time": "220.45",
"value": "strain"
},
{
"operation": "search",
"time": "231.882",
"value": "India-China relations"
},
{
"operation": "Now on",
"time": "233.253",
"value": "China–India relations"
},
{
"operation": "search",
"time": "241.116",
"value": "India-China border infrastructure"
},
{
"operation": "Now on",
"time": "242.125",
"value": "India-China border infrastructure"
},
{
"operation": "Highlighting",
"time": "258.612",
"value": "India-China border infrastructure_sentence_1"
},
{
"operation": "search",
"time": "272.789",
"value": "India-China border"
},
{
"operation": "Now on",
"time": "273.237",
"value": "India-China border"
},
{
"operation": "Page search",
"time": "285.993",
"value": "treaty"
},
{
"operation": "search",
"time": "297.684",
"value": "India-China relations"
},
{
"operation": "Now on",
"time": "298.34",
"value": "India-China relations"
},
{
"operation": "Page search",
"time": "300.531",
"value": "treaty"
},
{
"operation": "Highlighting",
"time": "360.73",
"value": "China–India relations_sentence_1"
},
{
"operation": "Now on",
"time": "376.98",
"value": "India-China relations"
},
{
"operation": "finish",
"time": "383.998",
"value": "finish"
}
]
}
|
15733
|
Thelyphonida or whip scorpions and vinegaroons have a body no longer than 50 mm and very large scorpion-like pedipalps or pincers.
|
{
"evidence": [
"Thelyphonida_sentence_19"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThelyphonida or whip scorpions and vinegaroons have a body no longer than 50 mm and very large scorpion-like pedipalps or pincers.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Thelyphonida_sentence_19"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7500,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.07",
"value": "?search="
},
{
"operation": "search",
"time": "26.038",
"value": "Thelyphonida"
},
{
"operation": "Now on",
"time": "27.066",
"value": "Thelyphonida"
},
{
"operation": "Highlighting",
"time": "30.94",
"value": "Thelyphonida_sentence_0"
},
{
"operation": "Highlighting",
"time": "42.235",
"value": "Thelyphonida_sentence_19"
},
{
"operation": "Highlighting",
"time": "74.748",
"value": "Thelyphonida_sentence_22"
},
{
"operation": "Highlighting deleted",
"time": "125.182",
"value": "Thelyphonida_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "130.424",
"value": "Thelyphonida_sentence_22"
},
{
"operation": "finish",
"time": "139.436",
"value": "finish"
}
]
}
|
11122
|
In 1900, the Tennessee Volunteers football team (coached by J.A. Pierce) tied with Vanderbilt with T 0-0 score on October 22.
|
{
"evidence": [
"Tennessee Volunteers football_sentence_0",
"History of Tennessee Volunteers football_sentence_1",
"History of Tennessee Volunteers football_sentence_19",
"J. A. Pierce_sentence_0",
"J. A. Pierce_sentence_1"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nIn 1900, the Tennessee Volunteers football team (coached by J.A. Pierce) tied with Vanderbilt with T 0-0 score on October 22.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Tennessee Volunteers football_sentence_0",
"History of Tennessee Volunteers football_sentence_1",
"History of Tennessee Volunteers football_sentence_19",
"J. A. Pierce_sentence_0",
"J. A. Pierce_sentence_1"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7501,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.706",
"value": "?search="
},
{
"operation": "search",
"time": "29.784",
"value": "Tennessee Volunteers football team"
},
{
"operation": "Now on",
"time": "31.882",
"value": "Tennessee Volunteers football"
},
{
"operation": "Highlighting",
"time": "42.178",
"value": "Tennessee Volunteers football_sentence_0"
},
{
"operation": "Page search",
"time": "61.122",
"value": "1900"
},
{
"operation": "hyperlink",
"time": "93.555",
"value": "Tennessee Volunteers football team"
},
{
"operation": "Now on",
"time": "94.89",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "96.804",
"value": "Tennessee Volunteers football"
},
{
"operation": "Now on",
"time": "97.84",
"value": "Tennessee Volunteers football"
},
{
"operation": "hyperlink",
"time": "113.36",
"value": "1 History Tennessee Volunteers football_section_0"
},
{
"operation": "hyperlink",
"time": "121.787",
"value": "History of Tennessee Volunteers football"
},
{
"operation": "Now on",
"time": "122.993",
"value": "History of Tennessee Volunteers football"
},
{
"operation": "Page search",
"time": "142.275",
"value": "1900"
},
{
"operation": "Page search",
"time": "174.979",
"value": "J.A. Pierce"
},
{
"operation": "Highlighting",
"time": "197.398",
"value": "History of Tennessee Volunteers football_sentence_1"
},
{
"operation": "hyperlink",
"time": "207.545",
"value": "1899"
},
{
"operation": "Now on",
"time": "208.761",
"value": "1899 Tennessee Volunteers football team"
},
{
"operation": "back-button-clicked",
"time": "219.304",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "219.507",
"value": "History of Tennessee Volunteers football"
},
{
"operation": "hyperlink",
"time": "252.644",
"value": "J. History of Tennessee Volunteers football_sentence_15\n\nA. History of Tennessee Volunteers football_sentence_16\n\nPierce"
},
{
"operation": "Now on",
"time": "253.768",
"value": "J. A. Pierce"
},
{
"operation": "Highlighting",
"time": "256.114",
"value": "J. A. Pierce_sentence_0"
},
{
"operation": "Highlighting",
"time": "259.243",
"value": "J. A. Pierce_sentence_1"
},
{
"operation": "back-button-clicked",
"time": "277.948",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "278.103",
"value": "History of Tennessee Volunteers football"
},
{
"operation": "Page search",
"time": "291.792",
"value": "Vanderbilt"
},
{
"operation": "Highlighting",
"time": "296.328",
"value": "History of Tennessee Volunteers football_sentence_19"
},
{
"operation": "hyperlink",
"time": "347.024",
"value": "Nash Buckingham"
},
{
"operation": "Now on",
"time": "348.061",
"value": "Nash Buckingham"
},
{
"operation": "back-button-clicked",
"time": "364.077",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "364.309",
"value": "History of Tennessee Volunteers football"
},
{
"operation": "finish",
"time": "427.568",
"value": "finish"
}
]
}
|
exfever_3669
|
American Horror Story is a Japanese anthology horror television series created by an American television writer, director, and producer.
|
{
"evidence": [
"American Horror Story",
"Ryan Murphy (writer)"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAmerican Horror Story is a Japanese anthology horror television series created by an American television writer, director, and producer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"American Horror Story",
"Ryan Murphy (writer)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7502,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
7434
|
Sophus Mads Jørgensen, a board member of Carlsberg Foundation is one of the founders of coordination chemistry who is still alive.
|
{
"evidence": [
"Sophus Mads Jørgensen_sentence_4"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nSophus Mads Jørgensen, a board member of Carlsberg Foundation is one of the founders of coordination chemistry who is still alive.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Sophus Mads Jørgensen_sentence_4"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7503,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.506",
"value": "?search="
},
{
"operation": "search",
"time": "17.564",
"value": "Sophus Mads Jørgensen"
},
{
"operation": "Now on",
"time": "18.738",
"value": "Sophus Mads Jørgensen"
},
{
"operation": "Highlighting",
"time": "30.13",
"value": "Sophus Mads Jørgensen_sentence_4"
},
{
"operation": "Highlighting",
"time": "37.079",
"value": "Sophus Mads Jørgensen_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "39.507",
"value": "Sophus Mads Jørgensen_sentence_0"
},
{
"operation": "finish",
"time": "153.479",
"value": "finish"
}
]
}
|
exfever_7496
|
Buddy Holly's style was influenced by country music, rhythm and blues acts, and a genre of Christian music.
|
{
"evidence": [
"Buddy Holly",
"Gospel music"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nBuddy Holly's style was influenced by country music, rhythm and blues acts, and a genre of Christian music.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Buddy Holly",
"Gospel music"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7504,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3889
|
Boston is the populous city of a political community founded for the common good in Connecticut.
|
{
"evidence": [
"Boston",
"Commonwealth (U.S. state)"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nBoston is the populous city of a political community founded for the common good in Connecticut.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Boston",
"Commonwealth (U.S. state)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7505,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
70916
|
Monte Westmore was Italian actor Rudolph Valentino's makeup artist until the actor's death in 1926 and helped Kurt Warner stop drinking.
|
{
"evidence": [
"Monte Westmore_sentence_10",
"Monte Westmore_sentence_11",
"Monte Westmore_sentence_14"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMonte Westmore was Italian actor Rudolph Valentino's makeup artist until the actor's death in 1926 and helped Kurt Warner stop drinking.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Monte Westmore_sentence_10",
"Monte Westmore_sentence_11",
"Monte Westmore_sentence_14"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7506,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.847",
"value": "?search="
},
{
"operation": "search",
"time": "8.751",
"value": "Monte Westmore"
},
{
"operation": "Now on",
"time": "9.776",
"value": "Monte Westmore"
},
{
"operation": "Highlighting",
"time": "185.82",
"value": "Monte Westmore_sentence_2"
},
{
"operation": "Highlighting",
"time": "515.71",
"value": "Monte Westmore_sentence_3"
},
{
"operation": "hyperlink",
"time": "524.846",
"value": "Gone with the Wind"
},
{
"operation": "Now on",
"time": "526.549",
"value": "Gone with the Wind (film)"
},
{
"operation": "back-button-clicked",
"time": "527.431",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "527.491",
"value": "Monte Westmore"
},
{
"operation": "Highlighting deleted",
"time": "528.786",
"value": "Monte Westmore_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "529.313",
"value": "Monte Westmore_sentence_3"
},
{
"operation": "Highlighting",
"time": "545.715",
"value": "Monte Westmore_sentence_10"
},
{
"operation": "Highlighting",
"time": "547.044",
"value": "Monte Westmore_sentence_11"
},
{
"operation": "Page search",
"time": "571.823",
"value": "kurt warner"
},
{
"operation": "hyperlink",
"time": "589.253",
"value": "H. Monte Westmore_sentence_12\n\nB. Warner"
},
{
"operation": "Now on",
"time": "589.88",
"value": "H. B. Warner"
},
{
"operation": "Page search",
"time": "597.103",
"value": "Monte Westmore"
},
{
"operation": "back-button-clicked",
"time": "606.898",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "606.967",
"value": "Monte Westmore"
},
{
"operation": "Highlighting",
"time": "609.267",
"value": "Monte Westmore_sentence_14"
},
{
"operation": "Highlighting",
"time": "610.09",
"value": "Monte Westmore_sentence_12Monte Westmore_sentence_13"
},
{
"operation": "Highlighting deleted",
"time": "613.217",
"value": "Monte Westmore_sentence_12Monte Westmore_sentence_13"
},
{
"operation": "hyperlink",
"time": "625.658",
"value": "H. Monte Westmore_sentence_12\n\nB. Warner"
},
{
"operation": "Now on",
"time": "626.059",
"value": "H. B. Warner"
},
{
"operation": "back-button-clicked",
"time": "635.676",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "635.74",
"value": "Monte Westmore"
},
{
"operation": "Highlighting deleted",
"time": "636.538",
"value": "Monte Westmore_sentence_11"
},
{
"operation": "Highlighting deleted",
"time": "637.2",
"value": "Monte Westmore_sentence_10"
},
{
"operation": "Highlighting deleted",
"time": "652.747",
"value": "Monte Westmore_sentence_14"
},
{
"operation": "Highlighting",
"time": "654.108",
"value": "Monte Westmore_sentence_10"
},
{
"operation": "Highlighting",
"time": "655.156",
"value": "Monte Westmore_sentence_11"
},
{
"operation": "Highlighting",
"time": "656.539",
"value": "Monte Westmore_sentence_14"
},
{
"operation": "finish",
"time": "660.869",
"value": "finish"
}
]
}
|
24295
|
Darlene McCoy started her career in music as a member of a quintet in 2005.
|
{
"evidence": [
"Darlene McCoy_sentence_1",
"Darlene McCoy_sentence_9"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nDarlene McCoy started her career in music as a member of a quintet in 2005.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Darlene McCoy_sentence_1",
"Darlene McCoy_sentence_9"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7507,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.034",
"value": "?search="
},
{
"operation": "search",
"time": "11.533",
"value": "Darlene McCoy"
},
{
"operation": "Now on",
"time": "12.233",
"value": "Darlene McCoy"
},
{
"operation": "Highlighting",
"time": "27.964",
"value": "Darlene McCoy_sentence_1"
},
{
"operation": "Highlighting",
"time": "75.449",
"value": "Darlene McCoy_sentence_9"
},
{
"operation": "finish",
"time": "92.345",
"value": "finish"
}
]
}
|
exfever_7047
|
The Times and its sister paper The Sunday Times, are published by Times Newspapers, since 1981 a subsidiary of News UK, in turn wholly owned by an American media and publishing company operating across digital real estate information, news media, book publishing, and cable television.
|
{
"evidence": [
"The Times",
"News Corp"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Times and its sister paper The Sunday Times, are published by Times Newspapers, since 1981 a subsidiary of News UK, in turn wholly owned by an American media and publishing company operating across digital real estate information, news media, book publishing, and cable television.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Times",
"News Corp"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7508,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_2883
|
Home for the Holidays is a 1995 film directed by Jodie Foster whose narrative follows a young woman who departs Chicago to spend her national holiday celebrated in the United Kingdom.
|
{
"evidence": [
"Home for the Holidays (1995 film)",
"Thanksgiving"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHome for the Holidays is a 1995 film directed by Jodie Foster whose narrative follows a young woman who departs Chicago to spend her national holiday celebrated in the United Kingdom.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Home for the Holidays (1995 film)",
"Thanksgiving"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7509,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
67004
|
The E1 domain contains two independent folding units, GFLD has a highly charged basic surface and is not packed on any organic substance.
|
{
"evidence": [
"APLP2_sentence_5",
"APLP2_sentence_6",
"APLP2_sentence_7"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe E1 domain contains two independent folding units, GFLD has a highly charged basic surface and is not packed on any organic substance.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"APLP2_sentence_5",
"APLP2_sentence_6",
"APLP2_sentence_7"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7510,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.733",
"value": "?search="
},
{
"operation": "search",
"time": "13.078",
"value": "Contains... E1 domain"
},
{
"operation": "Now on",
"time": "19.275",
"value": "undefined"
},
{
"operation": "search",
"time": "35.471",
"value": "Contains...GFLD"
},
{
"operation": "Now on",
"time": "36.832",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "60.853",
"value": "Amyloid precursor protein"
},
{
"operation": "Now on",
"time": "62.355",
"value": "Amyloid precursor protein"
},
{
"operation": "Page search",
"time": "81.079",
"value": "e1 domain"
},
{
"operation": "hyperlink",
"time": "137.565",
"value": "growth factor-like domain"
},
{
"operation": "Now on",
"time": "138.897",
"value": "Growth factor-like domain"
},
{
"operation": "hyperlink",
"time": "165.502",
"value": "protein domain"
},
{
"operation": "Now on",
"time": "175.527",
"value": "Protein domain"
},
{
"operation": "Page search",
"time": "215.844",
"value": "e1"
},
{
"operation": "search",
"time": "246.912",
"value": "Contains... E1 domain contains two independent folding units, GFLD has a highly charged basic surface"
},
{
"operation": "Now on",
"time": "250.603",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "255.879",
"value": "APLP2"
},
{
"operation": "Now on",
"time": "256.801",
"value": "APLP2"
},
{
"operation": "Highlighting",
"time": "264.437",
"value": "APLP2_sentence_5"
},
{
"operation": "Highlighting",
"time": "269.529",
"value": "APLP2_sentence_6"
},
{
"operation": "Page search",
"time": "293.435",
"value": "organic"
},
{
"operation": "search",
"time": "347.31",
"value": "gfld organic"
},
{
"operation": "Now on",
"time": "348.793",
"value": "undefined"
},
{
"operation": "search",
"time": "364.108",
"value": "gfld organic"
},
{
"operation": "Now on",
"time": "364.68",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "375.555",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "375.564",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "375.662",
"value": "APLP2"
},
{
"operation": "Page search",
"time": "391.1",
"value": "packed"
},
{
"operation": "hyperlink",
"time": "404.86",
"value": "beta-sheet"
},
{
"operation": "Now on",
"time": "405.716",
"value": "Beta sheet"
},
{
"operation": "back-button-clicked",
"time": "423.445",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "423.576",
"value": "APLP2"
},
{
"operation": "Highlighting",
"time": "433.996",
"value": "APLP2_sentence_7"
},
{
"operation": "finish",
"time": "445.881",
"value": "finish"
}
]
}
|
15944
|
Coleophora ornatipennella is a beetle that flies in May and June and does not prepare for hibernation.
|
{
"evidence": [
"Coleophora ornatipennella_sentence_0",
"Coleophora ornatipennella_sentence_5",
"Coleophora ornatipennella_sentence_8"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nColeophora ornatipennella is a beetle that flies in May and June and does not prepare for hibernation.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Coleophora ornatipennella_sentence_0",
"Coleophora ornatipennella_sentence_5",
"Coleophora ornatipennella_sentence_8"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7511,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.696",
"value": "?search="
},
{
"operation": "search",
"time": "13.2",
"value": "Coleophora ornatipennella"
},
{
"operation": "Now on",
"time": "14.029",
"value": "Coleophora ornatipennella"
},
{
"operation": "Highlighting",
"time": "17.579",
"value": "Coleophora ornatipennella_sentence_0"
},
{
"operation": "Highlighting",
"time": "28.527",
"value": "Coleophora ornatipennella_sentence_5"
},
{
"operation": "Highlighting",
"time": "33.096",
"value": "Coleophora ornatipennella_sentence_8"
},
{
"operation": "finish",
"time": "45.026",
"value": "finish"
}
]
}
|
exfever_525
|
Tropic Thunder is a 2008 action-comedy film directed by Ben Stiller and starring the American actor and producer who portrayed Iron Man in 2008.
|
{
"evidence": [
"Tropic Thunder",
"Robert Downey Jr."
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nTropic Thunder is a 2008 action-comedy film directed by Ben Stiller and starring the American actor and producer who portrayed Iron Man in 2008.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tropic Thunder",
"Robert Downey Jr."
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7512,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1801
|
Demi Lovato is an American singer-songwriter who signed with Hollywood Records and released her debut album that features primarily an R&B sound, with the lyrical content addressing other adolescent topics.
|
{
"evidence": [
"Demi Lovato",
"Don't Forget"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nDemi Lovato is an American singer-songwriter who signed with Hollywood Records and released her debut album that features primarily an R&B sound, with the lyrical content addressing other adolescent topics.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Demi Lovato",
"Don't Forget"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7513,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3672
|
Shahid Kapoor is an Indian actor who appears in the Indian Hindi-language film industry based in Mumbai.
|
{
"evidence": [
"Shahid Kapoor",
"Bollywood"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nShahid Kapoor is an Indian actor who appears in the Indian Hindi-language film industry based in Mumbai.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Shahid Kapoor",
"Bollywood"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7514,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3798
|
Leighton Meester has also appeared in films, including a 2010 American comedy film.
|
{
"evidence": [
"Leighton Meester",
"Country Strong"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nLeighton Meester has also appeared in films, including a 2010 American comedy film.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Leighton Meester",
"Country Strong"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7515,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5703
|
Carole King's first husband wrote with other composers including Michael Masser for a song that was initially recorded by American singer Thelma Houston.
|
{
"evidence": [
"Gerry Goffin",
"Theme from Mahogany (Do You Know Where You're Going To)"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCarole King's first husband wrote with other composers including Michael Masser for a song that was initially recorded by American singer Thelma Houston.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Gerry Goffin",
"Theme from Mahogany (Do You Know Where You're Going To)"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7516,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
50164
|
The First Division of the 1912 Liga Peruana de Football, which is the first season of top-flight Peruvian football, consists of eight teams such as Miraflores Sporting and Sport Alianza.
|
{
"evidence": [
"1912 Liga Peruana de Football_sentence_1"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe First Division of the 1912 Liga Peruana de Football, which is the first season of top-flight Peruvian football, consists of eight teams such as Miraflores Sporting and Sport Alianza.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"1912 Liga Peruana de Football_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7517,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.48",
"value": "undefined"
},
{
"operation": "search",
"time": "62.95",
"value": "Liga Peruana de Football"
},
{
"operation": "Now on",
"time": "65.161",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "90.343",
"value": "1913 Liga Peruana de Football"
},
{
"operation": "Now on",
"time": "91.515",
"value": "1913 Liga Peruana de Football"
},
{
"operation": "search",
"time": "114.616",
"value": "1912 Liga Peruana de Football"
},
{
"operation": "Now on",
"time": "115.513",
"value": "1912 Liga Peruana de Football"
},
{
"operation": "Highlighting",
"time": "148.26",
"value": "1912 Liga Peruana de Football_sentence_0"
},
{
"operation": "Highlighting",
"time": "401.7",
"value": "1912 Liga Peruana de Football_sentence_5"
},
{
"operation": "Highlighting",
"time": "410.296",
"value": "1912 Liga Peruana de Football_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "440.592",
"value": "1912 Liga Peruana de Football_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "492.651",
"value": "1912 Liga Peruana de Football_sentence_5"
},
{
"operation": "finish",
"time": "509.147",
"value": "finish"
}
]
}
|
78827
|
Shane Kirkwood, a South African rugby union player who regularly plays as a lock, made his Falcons debut against the Griffons which ended in a 30-all draw.
|
{
"evidence": [
"Shane Kirkwood_sentence_0",
"Shane Kirkwood_sentence_1",
"Shane Kirkwood_sentence_14",
"Shane Kirkwood_sentence_15"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nShane Kirkwood, a South African rugby union player who regularly plays as a lock, made his Falcons debut against the Griffons which ended in a 30-all draw.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Shane Kirkwood_sentence_0",
"Shane Kirkwood_sentence_1",
"Shane Kirkwood_sentence_14",
"Shane Kirkwood_sentence_15"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7518,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.573",
"value": "?search="
},
{
"operation": "search",
"time": "15.049",
"value": "Shane Kirkwood"
},
{
"operation": "Now on",
"time": "17.935",
"value": "Shane Kirkwood"
},
{
"operation": "Highlighting",
"time": "20.288",
"value": "Shane Kirkwood_sentence_0"
},
{
"operation": "Highlighting",
"time": "24.539",
"value": "Shane Kirkwood_sentence_1"
},
{
"operation": "Highlighting",
"time": "50.534",
"value": "Shane Kirkwood_sentence_7"
},
{
"operation": "Highlighting",
"time": "50.536",
"value": "Shane Kirkwood_section_1"
},
{
"operation": "Highlighting deleted",
"time": "60.219",
"value": "Shane Kirkwood_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "60.224",
"value": "Shane Kirkwood_section_1"
},
{
"operation": "Highlighting",
"time": "61.986",
"value": "Shane Kirkwood_sentence_14"
},
{
"operation": "Highlighting",
"time": "72.342",
"value": "Shane Kirkwood_sentence_15"
},
{
"operation": "finish",
"time": "89.808",
"value": "finish"
}
]
}
|
exfever_7818
|
The Great Debaters is a 2007 American biographical drama film directed by and starring an American actor, director, and producer who is known for his performances on the screen and on stage.
|
{
"evidence": [
"The Great Debaters",
"Denzel Washington"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Great Debaters is a 2007 American biographical drama film directed by and starring an American actor, director, and producer who is known for his performances on the screen and on stage.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Great Debaters",
"Denzel Washington"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7519,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_4309
|
Tanya Roberts was an American actress who play the role of Midge Pinciotti in an American television period sitcom that aired on Fox from August 23, 1998, to May 18, 2006.
|
{
"evidence": [
"Tanya Roberts",
"That '70s Show"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nTanya Roberts was an American actress who play the role of Midge Pinciotti in an American television period sitcom that aired on Fox from August 23, 1998, to May 18, 2006.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tanya Roberts",
"That '70s Show"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7520,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
72769
|
In UCAS (Universities and Colleges Admissions Service) Admission Statistics, there are 6,980 applicants in 2013 and 4,300 applicants in 2016.
|
{
"evidence": [
"University and college admission_sentence_275"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nIn UCAS (Universities and Colleges Admissions Service) Admission Statistics, there are 6,980 applicants in 2013 and 4,300 applicants in 2016.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"University and college admission_sentence_275"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7521,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.877",
"value": "?search="
},
{
"operation": "search",
"time": "23.429",
"value": "UCAS Admission Statistics"
},
{
"operation": "Now on",
"time": "25.159",
"value": "undefined"
},
{
"operation": "search",
"time": "46.028",
"value": "Universities and Colleges admissions service"
},
{
"operation": "Now on",
"time": "48.531",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "67.379",
"value": ".uk"
},
{
"operation": "Now on",
"time": "68.099",
"value": ".uk"
},
{
"operation": "Page search",
"time": "87.635",
"value": "universities and colleges admission service"
},
{
"operation": "back-button-clicked",
"time": "110.946",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "112.16",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "116.529",
"value": ".uk"
},
{
"operation": "Now on",
"time": "116.837",
"value": ".uk"
},
{
"operation": "Page search",
"time": "141.545",
"value": "support"
},
{
"operation": "back-button-clicked",
"time": "166.201",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "166.948",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "187.135",
"value": "Application process for universities and colleges UKPASS_section_3"
},
{
"operation": "Now on",
"time": "187.733",
"value": "UKPASS#Application process for universities and colleges UKPASS section 3"
},
{
"operation": "Highlighting",
"time": "209.855",
"value": "UKPASS_item_0_2"
},
{
"operation": "Highlighting",
"time": "209.856",
"value": "UKPASS_title"
},
{
"operation": "Highlighting",
"time": "209.856",
"value": "UKPASS_section_3"
},
{
"operation": "Highlighting deleted",
"time": "211.19",
"value": "UKPASS_item_0_2"
},
{
"operation": "Highlighting deleted",
"time": "211.19",
"value": "UKPASS_title"
},
{
"operation": "Highlighting deleted",
"time": "211.191",
"value": "UKPASS_section_3"
},
{
"operation": "back-button-clicked",
"time": "213.174",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "213.825",
"value": "undefined"
},
{
"operation": "search",
"time": "236.791",
"value": "2016 UCAS Admission Statistics"
},
{
"operation": "Now on",
"time": "237.513",
"value": "undefined"
},
{
"operation": "search",
"time": "259.661",
"value": "UCAS"
},
{
"operation": "Now on",
"time": "260.355",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "271.821",
"value": "University and college admission"
},
{
"operation": "Now on",
"time": "273.192",
"value": "University and college admission"
},
{
"operation": "hyperlink",
"time": "298.684",
"value": "31.1 Undergraduate entry University and college admission_section_51"
},
{
"operation": "Highlighting",
"time": "332.444",
"value": "University and college admission_sentence_275"
},
{
"operation": "Highlighting deleted",
"time": "333.379",
"value": "University and college admission_sentence_275"
},
{
"operation": "Highlighting",
"time": "335.259",
"value": "University and college admission_sentence_275"
},
{
"operation": "back-button-clicked",
"time": "345.163",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "345.852",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "356.867",
"value": "UCAS clearing Keele University_section_32"
},
{
"operation": "Now on",
"time": "359.597",
"value": "Keele University#UCAS clearing Keele University section 32"
},
{
"operation": "search",
"time": "377.85",
"value": "UCAS 2016"
},
{
"operation": "Now on",
"time": "378.757",
"value": "undefined"
},
{
"operation": "search",
"time": "409.848",
"value": "6,980 applicants"
},
{
"operation": "Now on",
"time": "411.461",
"value": "undefined"
},
{
"operation": "search",
"time": "436.775",
"value": "UCAS Admission Statistics 2016"
},
{
"operation": "Now on",
"time": "437.329",
"value": "undefined"
},
{
"operation": "search",
"time": "460.814",
"value": "UCAS clearning process"
},
{
"operation": "Now on",
"time": "461.748",
"value": "undefined"
},
{
"operation": "search",
"time": "477.582",
"value": "Undergraduate admissions UK"
},
{
"operation": "Now on",
"time": "478.584",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "494.181",
"value": "next 20"
},
{
"operation": "Now on",
"time": "494.964",
"value": "undefined"
},
{
"operation": "finish",
"time": "508.716",
"value": "finish"
}
]
}
|
exfever_6687
|
Tamera Mowry is an Australian actress, television personality, and former singer who starred in a 2005 Disney Channel original movie produced by Broomsticks Productions Limited.
|
{
"evidence": [
"Tamera Mowry",
"Twitches (film)"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nTamera Mowry is an Australian actress, television personality, and former singer who starred in a 2005 Disney Channel original movie produced by Broomsticks Productions Limited.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tamera Mowry",
"Twitches (film)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7522,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5791
|
Sir Michael Caine is an English actor who stars in a British comedy-drama film directed by Lewis Gilbert.
|
{
"evidence": [
"Michael Caine",
"Alfie (1966 film)"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nSir Michael Caine is an English actor who stars in a British comedy-drama film directed by Lewis Gilbert.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Michael Caine",
"Alfie (1966 film)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7523,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3937
|
Maya Khabira Rudolph is an American actress who has played supporting roles in films, including a 2006 American science fiction comedy film that stars an American actor, comedian, writer, director, and podcast host.
|
{
"evidence": [
"Maya Rudolph",
"Idiocracy",
"Dax Shepard"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMaya Khabira Rudolph is an American actress who has played supporting roles in films, including a 2006 American science fiction comedy film that stars an American actor, comedian, writer, director, and podcast host.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Maya Rudolph",
"Idiocracy",
"Dax Shepard"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7524,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1664
|
The Twilight Saga is a series of four vampire-themed rom-com novels written by Stephanie Meyer that follow Bella Swan, a girl who moves to a city in southwest Clallam County, which is a state in the Pacific Northwest region of the Western United States.
|
{
"evidence": [
"Twilight (novel series)",
"Forks, Washington",
"Washington (state)"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Twilight Saga is a series of four vampire-themed rom-com novels written by Stephanie Meyer that follow Bella Swan, a girl who moves to a city in southwest Clallam County, which is a state in the Pacific Northwest region of the Western United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Twilight (novel series)",
"Forks, Washington",
"Washington (state)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7525,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8588
|
The Adjustment Bureau is the last American film directed by George Nolfi.
|
{
"evidence": [
"The Adjustment Bureau",
"Science fiction",
"Technology"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Adjustment Bureau is the last American film directed by George Nolfi.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Adjustment Bureau",
"Science fiction",
"Technology"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7526,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
82793
|
Halocynthia papillosa is a sea peach or sea turtle in the genus of Halocynthia, in the Pyuridae family, within the Pleurogona order.
|
{
"evidence": [
"Halocynthia papillosa_sentence_1"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHalocynthia papillosa is a sea peach or sea turtle in the genus of Halocynthia, in the Pyuridae family, within the Pleurogona order.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Halocynthia papillosa_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7527,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.034",
"value": "?search="
},
{
"operation": "search",
"time": "8.117",
"value": "Halocynthia papillosa"
},
{
"operation": "Now on",
"time": "9.049",
"value": "Halocynthia papillosa"
},
{
"operation": "Highlighting",
"time": "12.015",
"value": "Halocynthia papillosa_sentence_1"
},
{
"operation": "finish",
"time": "66.366",
"value": "finish"
}
]
}
|
51494
|
Landbay is a company founded by John Goodall and Gray Stern that facilitates peer-to-peer lending platform for residential buy-to-let mortgages.
|
{
"evidence": [
"Landbay_sentence_0",
"Landbay_sentence_5"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nLandbay is a company founded by John Goodall and Gray Stern that facilitates peer-to-peer lending platform for residential buy-to-let mortgages.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Landbay_sentence_0",
"Landbay_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7528,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.436",
"value": "?search="
},
{
"operation": "search",
"time": "9.637",
"value": "Landbay"
},
{
"operation": "Now on",
"time": "10.694",
"value": "Landbay"
},
{
"operation": "Highlighting",
"time": "23.049",
"value": "Landbay_sentence_0"
},
{
"operation": "Highlighting",
"time": "30.501",
"value": "Landbay_sentence_5"
},
{
"operation": "finish",
"time": "36.348",
"value": "finish"
}
]
}
|
exfever_2569
|
Lauren Bacall was an English actress who married an American film and stage actor.
|
{
"evidence": [
"Lauren Bacall",
"Humphrey Bogart"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nLauren Bacall was an English actress who married an American film and stage actor.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Lauren Bacall",
"Humphrey Bogart"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7529,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3242
|
Hamlet is a 1990 drama film directed by Franco Zeffirelli and starring talented artists, including an English actor who came to prominence in the 1960s.
|
{
"evidence": [
"Hamlet (1990 film)",
"Alan Bates"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHamlet is a 1990 drama film directed by Franco Zeffirelli and starring talented artists, including an English actor who came to prominence in the 1960s.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Hamlet (1990 film)",
"Alan Bates"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7530,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
52014
|
At the age of ten, Catherine Heigel played Cosette in a radio adaptation of Les Misérables, a French historical novel by Victor Hugo, published in 1862, and considered as one of the greatest novels of the 19th century.
|
{
"evidence": [
"Catherine Hiegel_sentence_4",
"Les Misérables_sentence_3"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAt the age of ten, Catherine Heigel played Cosette in a radio adaptation of Les Misérables, a French historical novel by Victor Hugo, published in 1862, and considered as one of the greatest novels of the 19th century.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Catherine Hiegel_sentence_4",
"Les Misérables_sentence_3"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7531,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.723",
"value": "?search="
},
{
"operation": "search",
"time": "12.641",
"value": "Contains...Catherine Heigel"
},
{
"operation": "Now on",
"time": "13.247",
"value": "undefined"
},
{
"operation": "search",
"time": "45.688",
"value": "Les Misérables (disambiguation)"
},
{
"operation": "Now on",
"time": "46.63",
"value": "Les Misérables (disambiguation)"
},
{
"operation": "hyperlink",
"time": "61.489",
"value": "Les Misérables (radio series)"
},
{
"operation": "Now on",
"time": "62.142",
"value": "Les Misérables (radio series)"
},
{
"operation": "Page search",
"time": "110.704",
"value": "CATHERINE"
},
{
"operation": "search",
"time": "143.174",
"value": "Contains... Catherine Heigel"
},
{
"operation": "Now on",
"time": "143.904",
"value": "undefined"
},
{
"operation": "search",
"time": "170.286",
"value": "Les Misérables CATHERINE HEIGEL"
},
{
"operation": "Now on",
"time": "170.922",
"value": "undefined"
},
{
"operation": "search",
"time": "180.862",
"value": "Les Misérables (disambiguation)"
},
{
"operation": "Now on",
"time": "181.4",
"value": "Les Misérables (disambiguation)"
},
{
"operation": "hyperlink",
"time": "202.924",
"value": "Los miserables (disambiguation)"
},
{
"operation": "Now on",
"time": "203.627",
"value": "Los miserables"
},
{
"operation": "back-button-clicked",
"time": "208.968",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "209.051",
"value": "Les Misérables (disambiguation)"
},
{
"operation": "search",
"time": "223.014",
"value": "CATHERINE COSETTE"
},
{
"operation": "Now on",
"time": "223.864",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "235.553",
"value": "Catherine Hiegel"
},
{
"operation": "Now on",
"time": "239.43",
"value": "Catherine Hiegel"
},
{
"operation": "Highlighting",
"time": "241.49",
"value": "Catherine Hiegel_sentence_4"
},
{
"operation": "hyperlink",
"time": "246.238",
"value": "Les Misérables"
},
{
"operation": "Now on",
"time": "256.917",
"value": "Les Misérables (musical)"
},
{
"operation": "hyperlink",
"time": "260.332",
"value": "novel of the same name"
},
{
"operation": "Now on",
"time": "261.815",
"value": "Les Misérables"
},
{
"operation": "Highlighting",
"time": "269.239",
"value": "Les Misérables_sentence_3"
},
{
"operation": "finish",
"time": "289.379",
"value": "finish"
}
]
}
|
exfever_1031
|
Leicester City Football Club is a professional football club that competes in the Premier League, the highest level of a series of interconnected leagues for men's association football clubs in a country that is part of the United Kingdom.
|
{
"evidence": [
"Leicester City F.C.",
"English football league system",
"England"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nLeicester City Football Club is a professional football club that competes in the Premier League, the highest level of a series of interconnected leagues for men's association football clubs in a country that is part of the United Kingdom.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Leicester City F.C.",
"English football league system",
"England"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7532,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
31045
|
Hextall, a landowner born in 1861 neither at Canonbury House nor in Islington, received title on August 4, 1910, then quickly saw the possibility of developing his ranch into a garden suburb for the wealthy.
|
{
"evidence": [
"John Hextall_sentence_1"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHextall, a landowner born in 1861 neither at Canonbury House nor in Islington, received title on August 4, 1910, then quickly saw the possibility of developing his ranch into a garden suburb for the wealthy.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"John Hextall_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7533,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.522",
"value": "?search="
},
{
"operation": "search",
"time": "19.842",
"value": "hextall"
},
{
"operation": "Now on",
"time": "20.359",
"value": "Hextall"
},
{
"operation": "hyperlink",
"time": "29.85",
"value": "John Hextall"
},
{
"operation": "Now on",
"time": "30.232",
"value": "John Hextall"
},
{
"operation": "Highlighting",
"time": "36.002",
"value": "John Hextall_sentence_1"
},
{
"operation": "finish",
"time": "61.281",
"value": "finish"
}
]
}
|
exfever_1356
|
Mozilla Firefox is an expensive and open-source web browser developed by an American non-profit organization that exists to support and collectively lead the open-source Mozilla project.
|
{
"evidence": [
"Firefox",
"Mozilla Foundation"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMozilla Firefox is an expensive and open-source web browser developed by an American non-profit organization that exists to support and collectively lead the open-source Mozilla project.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Firefox",
"Mozilla Foundation"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7534,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
82265
|
The Mo was originally a house named after Morag, born 1881, daughter of Sir Thomas Digby Pigott a member of the Upcher family; and Richard Frederick Hayward, M.C., K.C., a maritime lawyer and farmer born on March 1879, was its last owner.
|
{
"evidence": [
"The Mo Sheringham Museum_sentence_5",
"The Mo Sheringham Museum_sentence_6"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Mo was originally a house named after Morag, born 1881, daughter of Sir Thomas Digby Pigott a member of the Upcher family; and Richard Frederick Hayward, M.C., K.C., a maritime lawyer and farmer born on March 1879, was its last owner.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"The Mo Sheringham Museum_sentence_5",
"The Mo Sheringham Museum_sentence_6"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7535,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.636",
"value": "?search="
},
{
"operation": "search",
"time": "9.162",
"value": "The Mo"
},
{
"operation": "Now on",
"time": "9.624",
"value": "undefined"
},
{
"operation": "search",
"time": "28.369",
"value": "\"The Mo was originally a house named after Morag, born 1881\""
},
{
"operation": "Now on",
"time": "28.839",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "32.602",
"value": "The Mo Sheringham Museum"
},
{
"operation": "Now on",
"time": "33.004",
"value": "The Mo Sheringham Museum"
},
{
"operation": "Highlighting",
"time": "46.051",
"value": "The Mo Sheringham Museum_sentence_5"
},
{
"operation": "Highlighting",
"time": "88.749",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting deleted",
"time": "118.215",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting",
"time": "118.912",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting deleted",
"time": "121.355",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting",
"time": "122.813",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "search",
"time": "126.388",
"value": "Richard Frederick Hayward"
},
{
"operation": "Now on",
"time": "128.083",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "132.133",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "132.207",
"value": "The Mo Sheringham Museum"
},
{
"operation": "Highlighting",
"time": "183.415",
"value": "The Mo Sheringham Museum_sentence_4"
},
{
"operation": "Highlighting",
"time": "183.416",
"value": "The Mo Sheringham Museum_section_0"
},
{
"operation": "Highlighting deleted",
"time": "185.022",
"value": "The Mo Sheringham Museum_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "185.024",
"value": "The Mo Sheringham Museum_section_0"
},
{
"operation": "Highlighting",
"time": "185.357",
"value": "The Mo Sheringham Museum_sentence_4"
},
{
"operation": "Highlighting",
"time": "185.358",
"value": "The Mo Sheringham Museum_section_0"
},
{
"operation": "Highlighting deleted",
"time": "186.844",
"value": "The Mo Sheringham Museum_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "186.846",
"value": "The Mo Sheringham Museum_section_0"
},
{
"operation": "Highlighting deleted",
"time": "187.936",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting",
"time": "191.063",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting deleted",
"time": "192.64",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "Highlighting",
"time": "195.788",
"value": "The Mo Sheringham Museum_sentence_6"
},
{
"operation": "search",
"time": "279.798",
"value": "Richard Frederick Hayward, M.C., K.C., a maritime lawyer and farmer"
},
{
"operation": "Now on",
"time": "281.358",
"value": "undefined"
},
{
"operation": "search",
"time": "305.243",
"value": "Richard Frederick Hayward"
},
{
"operation": "Now on",
"time": "305.949",
"value": "undefined"
},
{
"operation": "finish",
"time": "379.596",
"value": "finish"
}
]
}
|
exfever_5814
|
Paul Thomas Anderson is an American film director whose fifth film is the 2007 American period drama film with a large-scale, sweeping scope, and spectacle style of filmmaking.
|
{
"evidence": [
"Paul Thomas Anderson",
"There Will Be Blood",
"Epic film"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nPaul Thomas Anderson is an American film director whose fifth film is the 2007 American period drama film with a large-scale, sweeping scope, and spectacle style of filmmaking.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Paul Thomas Anderson",
"There Will Be Blood",
"Epic film"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7536,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
70575
|
5 Schoolhouse is a historic two-room district schoolhouse at Intervale and Day Mountain Roads in Temple, Maine built in 1810–11.
|
{
"evidence": [
"Temple Intervale School_sentence_1"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\n5 Schoolhouse is a historic two-room district schoolhouse at Intervale and Day Mountain Roads in Temple, Maine built in 1810–11.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Temple Intervale School_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7537,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.742",
"value": "?search="
},
{
"operation": "search",
"time": "10.953",
"value": "Contains...5 Schoolhouse"
},
{
"operation": "Now on",
"time": "14.316",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "22.816",
"value": "Bingham District No. 5 Schoolhouse"
},
{
"operation": "Now on",
"time": "23.711",
"value": "Bingham District No. 5 Schoolhouse"
},
{
"operation": "Highlighting",
"time": "48.504",
"value": "Bingham District No. 5 Schoolhouse_sentence_1"
},
{
"operation": "Highlighting",
"time": "48.505",
"value": "Bingham District No. 5 Schoolhouse_title"
},
{
"operation": "back-button-clicked",
"time": "55.566",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "56.359",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "62.713",
"value": "Mason District Number 5 Schoolhouse"
},
{
"operation": "Now on",
"time": "63.247",
"value": "Mason District Number 5 Schoolhouse"
},
{
"operation": "Highlighting",
"time": "87.937",
"value": "Mason District Number 5 Schoolhouse_sentence_0"
},
{
"operation": "Highlighting",
"time": "87.938",
"value": "Mason District Number 5 Schoolhouse_title"
},
{
"operation": "back-button-clicked",
"time": "99.792",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "100.915",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "110.758",
"value": "District No. 5 Schoolhouse"
},
{
"operation": "Now on",
"time": "112.208",
"value": "District No. 5 Schoolhouse"
},
{
"operation": "search",
"time": "135.231",
"value": "Contains...at Intervale and Day Mountain Roads in Temple, Maine"
},
{
"operation": "Now on",
"time": "136.532",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "140.932",
"value": "Temple Intervale School"
},
{
"operation": "Now on",
"time": "141.839",
"value": "Temple Intervale School"
},
{
"operation": "Highlighting",
"time": "147.413",
"value": "Temple Intervale School_sentence_1"
},
{
"operation": "finish",
"time": "169.355",
"value": "finish"
}
]
}
|
exfever_3754
|
Sir Timothy Miles Bindon Rice is an English lyricist and author who wrote a sung-through musical comedy that is based on the character of Joseph from the Bible's Book of Genesis.
|
{
"evidence": [
"Tim Rice",
"Joseph and the Amazing Technicolor Dreamcoat"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nSir Timothy Miles Bindon Rice is an English lyricist and author who wrote a sung-through musical comedy that is based on the character of Joseph from the Bible's Book of Genesis.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tim Rice",
"Joseph and the Amazing Technicolor Dreamcoat"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7538,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
26418
|
Julie Dubela born January 16, 1991) is a German singer from Stratham, New Hampshire, Dubela first appeared on national TV in 2003, when she participated in Fox TV's American Juniors, a reality TV competition, to select members for a new band.
|
{
"evidence": [
"Julie Dubela_sentence_0"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJulie Dubela born January 16, 1991) is a German singer from Stratham, New Hampshire, Dubela first appeared on national TV in 2003, when she participated in Fox TV's American Juniors, a reality TV competition, to select members for a new band.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Julie Dubela_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7539,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.72",
"value": "?search="
},
{
"operation": "search",
"time": "38.339",
"value": "Julie Dubela"
},
{
"operation": "Now on",
"time": "39.017",
"value": "Julie Dubela"
},
{
"operation": "Highlighting",
"time": "65.233",
"value": "Julie Dubela_sentence_0"
},
{
"operation": "finish",
"time": "75.751",
"value": "finish"
}
]
}
|
83458
|
A65 autoroute, also known as Autoroute de Gascogne, is approximately 150 kilometres (93 mi), and was built by a joint venture A'LIENOR.
|
{
"evidence": [
"A65 autoroute_sentence_3",
"A65 autoroute_sentence_5"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nA65 autoroute, also known as Autoroute de Gascogne, is approximately 150 kilometres (93 mi), and was built by a joint venture A'LIENOR.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"A65 autoroute_sentence_3",
"A65 autoroute_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7540,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.093",
"value": "?search="
},
{
"operation": "search",
"time": "14.511",
"value": "Contains...A65 autoroute"
},
{
"operation": "Now on",
"time": "15.779",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "17.608",
"value": "A65 autoroute"
},
{
"operation": "Now on",
"time": "18.072",
"value": "A65 autoroute"
},
{
"operation": "Highlighting",
"time": "26.679",
"value": "A65 autoroute_sentence_3"
},
{
"operation": "Highlighting",
"time": "37.304",
"value": "A65 autoroute_sentence_5"
},
{
"operation": "finish",
"time": "44.541",
"value": "finish"
}
]
}
|
23398
|
The Metaphysical Society of America was sued in 1950 by Paul Weiss.
|
{
"evidence": [
"Metaphysical Society of America_sentence_0"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Metaphysical Society of America was sued in 1950 by Paul Weiss.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Metaphysical Society of America_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7541,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.578",
"value": "?search="
},
{
"operation": "search",
"time": "12.041",
"value": "Metaphysical Society of America"
},
{
"operation": "Now on",
"time": "12.958",
"value": "Metaphysical Society of America"
},
{
"operation": "Page search",
"time": "18.32",
"value": "sue"
},
{
"operation": "Page search",
"time": "27.737",
"value": "suit"
},
{
"operation": "Highlighting",
"time": "34.951",
"value": "Metaphysical Society of America_sentence_0"
},
{
"operation": "finish",
"time": "66.427",
"value": "finish"
}
]
}
|
exfever_2202
|
Inside Man is a 2009 American heist thriller film starring an artist who is known for his frequent collaborations with an English film director, producer, actor, screenwriter, and editor.
|
{
"evidence": [
"Inside Man",
"Denzel Washington",
"Tony Scott"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nInside Man is a 2009 American heist thriller film starring an artist who is known for his frequent collaborations with an English film director, producer, actor, screenwriter, and editor.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Inside Man",
"Denzel Washington",
"Tony Scott"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7542,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
36415
|
MFC "Ayat", a futsal club based in Rudny, was created in 2007 following the results of Championship of Republic of Kazakhstan on a futsal among the first league teams; in which teams from all Republic took part the "Ayat" team wins first place, and receives the permit in the top division.
|
{
"evidence": [
"MFC Ayat_sentence_11",
"MFC Ayat_sentence_12"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMFC \"Ayat\", a futsal club based in Rudny, was created in 2007 following the results of Championship of Republic of Kazakhstan on a futsal among the first league teams; in which teams from all Republic took part the \"Ayat\" team wins first place, and receives the permit in the top division.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"MFC Ayat_sentence_11",
"MFC Ayat_sentence_12"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7543,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.205",
"value": "?search="
},
{
"operation": "search",
"time": "11.928",
"value": "MFC \"Ayat\""
},
{
"operation": "Now on",
"time": "12.64",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "124.441",
"value": "MFC Ayat"
},
{
"operation": "Now on",
"time": "125.226",
"value": "MFC Ayat"
},
{
"operation": "Highlighting",
"time": "133.753",
"value": "MFC Ayat_sentence_0"
},
{
"operation": "Highlighting",
"time": "135.505",
"value": "MFC Ayat_sentence_1"
},
{
"operation": "Highlighting",
"time": "151.881",
"value": "MFC Ayat_sentence_11"
},
{
"operation": "Highlighting deleted",
"time": "155.313",
"value": "MFC Ayat_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "155.801",
"value": "MFC Ayat_sentence_1"
},
{
"operation": "Highlighting",
"time": "190.425",
"value": "MFC Ayat_sentence_12"
},
{
"operation": "finish",
"time": "206.504",
"value": "finish"
}
]
}
|
exfever_2221
|
When blending into the society outside of her homeland, Wonder Woman sometimes adopts her alter ego, which is not known to the general populace, Diana Prince.
|
{
"evidence": [
"Wonder Woman",
"Secret identity"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nWhen blending into the society outside of her homeland, Wonder Woman sometimes adopts her alter ego, which is not known to the general populace, Diana Prince.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Wonder Woman",
"Secret identity"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7544,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
84859
|
Methane is a primary component of natural gas, and its process inevitably releases a significant amount of methane into the atmosphere, and according to the EPA Inventory of U.S Greenhouse Gas Emissions and Sinks: 1990–2015 report, 2015 methane emissions from natural gas and petroleum systems totaled 6.5 Tg per year in the United States, and has a 20-year global warming potential of methane is 84, that is, over a 20-year period, it traps 84 times more heat per mass unit than carbon dioxide (CO2) and 32 times the effect when accounting for aerosol interactions.
|
{
"evidence": [
"Atmospheric methane_sentence_121"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMethane is a primary component of natural gas, and its process inevitably releases a significant amount of methane into the atmosphere, and according to the EPA Inventory of U.S Greenhouse Gas Emissions and Sinks: 1990–2015 report, 2015 methane emissions from natural gas and petroleum systems totaled 6.5 Tg per year in the United States, and has a 20-year global warming potential of methane is 84, that is, over a 20-year period, it traps 84 times more heat per mass unit than carbon dioxide (CO2) and 32 times the effect when accounting for aerosol interactions.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Atmospheric methane_sentence_121"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7545,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.446",
"value": "?search="
},
{
"operation": "search",
"time": "14.471",
"value": "process inevitably releases a significant amount of methane"
},
{
"operation": "Now on",
"time": "15.817",
"value": "undefined"
},
{
"operation": "search",
"time": "36.278",
"value": "significant amount of methane into the atmosphere"
},
{
"operation": "Now on",
"time": "37.907",
"value": "undefined"
},
{
"operation": "search",
"time": "54.661",
"value": "20-year global warming potential of methane is 84"
},
{
"operation": "Now on",
"time": "57.286",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "61.859",
"value": "Release of stored arctic methane due to global warming Atmospheric methane_section_35"
},
{
"operation": "Now on",
"time": "63.74",
"value": "Atmospheric methane#Release of stored arctic methane due to global warming Atmospheric methane section 35"
},
{
"operation": "Page search",
"time": "76.679",
"value": "20-year global warming potential of methane is 84"
},
{
"operation": "Highlighting",
"time": "108.662",
"value": "Atmospheric methane_sentence_3"
},
{
"operation": "Page search",
"time": "114.885",
"value": "primary"
},
{
"operation": "Highlighting",
"time": "126.564",
"value": "Atmospheric methane_sentence_120"
},
{
"operation": "Highlighting",
"time": "151.993",
"value": "Atmospheric methane_sentence_121"
},
{
"operation": "finish",
"time": "249.468",
"value": "finish"
}
]
}
|
exfever_3758
|
Frederick Seibert is a television producer who has held leading positions with a company based in New York City since March 2007.
|
{
"evidence": [
"Fred Seibert",
"YouTube Next Lab and Audience Development Group"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nFrederick Seibert is a television producer who has held leading positions with a company based in New York City since March 2007.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Fred Seibert",
"YouTube Next Lab and Audience Development Group"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7546,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
88123
|
Omar Abdillahi Charmarke competed in four local marathon competitions from 1985 to 1992.
|
{
"evidence": [
"Omar Abdillahi Charmarke_sentence_1"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nOmar Abdillahi Charmarke competed in four local marathon competitions from 1985 to 1992.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Omar Abdillahi Charmarke_sentence_1"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7547,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.01",
"value": "?search="
},
{
"operation": "search",
"time": "3.998",
"value": "Omar Abdillahi Charmarke"
},
{
"operation": "Now on",
"time": "4.589",
"value": "Omar Abdillahi Charmarke"
},
{
"operation": "Highlighting",
"time": "26.037",
"value": "Omar Abdillahi Charmarke_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "31.153",
"value": "Omar Abdillahi Charmarke_sentence_0"
},
{
"operation": "Highlighting",
"time": "31.795",
"value": "Omar Abdillahi Charmarke_sentence_1"
},
{
"operation": "finish",
"time": "43.31",
"value": "finish"
}
]
}
|
exfever_7428
|
Despicable Me 3 is a 2017 computer-animated film that emphasizes humor, designed to make the audience laugh through amusement.
|
{
"evidence": [
"Despicable Me 3",
"Comedy film"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nDespicable Me 3 is a 2017 computer-animated film that emphasizes humor, designed to make the audience laugh through amusement.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Despicable Me 3",
"Comedy film"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7548,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
69584
|
Democrat candidate John Patrick Hopkins won the Chicago mayoral special election of 1893 with a narrow margin of victory against Republican George Bell Swift following the assassination of Mayor Carter Harrison Sr.
|
{
"evidence": [
"John Patrick Hopkins_sentence_14"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nDemocrat candidate John Patrick Hopkins won the Chicago mayoral special election of 1893 with a narrow margin of victory against Republican George Bell Swift following the assassination of Mayor Carter Harrison Sr.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"John Patrick Hopkins_sentence_14"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7549,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.169",
"value": "?search="
},
{
"operation": "Now on",
"time": "6.207",
"value": "?search="
},
{
"operation": "search",
"time": "13.305",
"value": "John Patrick Hopkins"
},
{
"operation": "Now on",
"time": "14.277",
"value": "John Patrick Hopkins"
},
{
"operation": "Highlighting",
"time": "58.422",
"value": "John Patrick Hopkins_sentence_14"
},
{
"operation": "hyperlink",
"time": "123.307",
"value": "Mayor of Chicago"
},
{
"operation": "Now on",
"time": "124.467",
"value": "Mayor of Chicago"
},
{
"operation": "Highlighting",
"time": "163.754",
"value": "Mayor of Chicago_cell_2_36_2"
},
{
"operation": "Highlighting",
"time": "163.757",
"value": "Mayor of Chicago_title"
},
{
"operation": "Highlighting",
"time": "163.762",
"value": "Mayor of Chicago_section_3"
},
{
"operation": "Highlighting",
"time": "163.765",
"value": "Mayor of Chicago_section_5"
},
{
"operation": "Highlighting",
"time": "163.789",
"value": "Mayor of Chicago_header_cell_2_0_2"
},
{
"operation": "Highlighting deleted",
"time": "193.906",
"value": "Mayor of Chicago_cell_2_36_2"
},
{
"operation": "Highlighting deleted",
"time": "193.913",
"value": "Mayor of Chicago_title"
},
{
"operation": "Highlighting deleted",
"time": "193.919",
"value": "Mayor of Chicago_section_3"
},
{
"operation": "Highlighting deleted",
"time": "193.926",
"value": "Mayor of Chicago_section_5"
},
{
"operation": "Highlighting deleted",
"time": "193.951",
"value": "Mayor of Chicago_header_cell_2_0_2"
},
{
"operation": "hyperlink",
"time": "196.576",
"value": "George Bell Swift"
},
{
"operation": "Now on",
"time": "198.548",
"value": "George Bell Swift"
},
{
"operation": "finish",
"time": "245.287",
"value": "finish"
}
]
}
|
21500
|
Twelve International Association of Dental Students Mid-year meetings starting from 1942-2021 were held in different countries, the first meeting in 1942 was held in Bangkok Thailand, while the last four meetings were done online.
|
{
"evidence": [
"International Association of Dental Students_sentence_1"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nTwelve International Association of Dental Students Mid-year meetings starting from 1942-2021 were held in different countries, the first meeting in 1942 was held in Bangkok Thailand, while the last four meetings were done online.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"International Association of Dental Students_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7550,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.972",
"value": "undefined"
},
{
"operation": "search",
"time": "43.224",
"value": "Twelve International Association of Dental Students"
},
{
"operation": "Now on",
"time": "44.075",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "46.568",
"value": "International partnerships University of Pennsylvania_section_18"
},
{
"operation": "Now on",
"time": "48.065",
"value": "University of Pennsylvania#International partnerships University of Pennsylvania section 18"
},
{
"operation": "Page search",
"time": "81.354",
"value": "Twelve"
},
{
"operation": "Highlighting",
"time": "128.284",
"value": "University of Pennsylvania_sentence_239"
},
{
"operation": "Highlighting",
"time": "128.285",
"value": "University of Pennsylvania_title"
},
{
"operation": "Highlighting",
"time": "128.286",
"value": "University of Pennsylvania_section_22"
},
{
"operation": "Highlighting",
"time": "128.287",
"value": "University of Pennsylvania_section_23"
},
{
"operation": "Page search",
"time": "135.951",
"value": "Twelve International"
},
{
"operation": "back-button-clicked",
"time": "168.68",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "169.558",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "210.427",
"value": "Boston University International Affairs Association"
},
{
"operation": "Now on",
"time": "212.289",
"value": "Boston University International Affairs Association"
},
{
"operation": "Page search",
"time": "216.874",
"value": "Twelve International"
},
{
"operation": "search",
"time": "262.174",
"value": "International Association of Dental Students"
},
{
"operation": "Now on",
"time": "263.644",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "265.789",
"value": "International Association of Dental Students"
},
{
"operation": "Now on",
"time": "266.71",
"value": "International Association of Dental Students"
},
{
"operation": "Page search",
"time": "295.725",
"value": "mid"
},
{
"operation": "Page search",
"time": "324.509",
"value": "mid year"
},
{
"operation": "Page search",
"time": "389.605",
"value": "1942"
},
{
"operation": "Page search",
"time": "389.892",
"value": "1942"
},
{
"operation": "Highlighting",
"time": "418.266",
"value": "International Association of Dental Students_sentence_0"
},
{
"operation": "Highlighting",
"time": "456.705",
"value": "International Association of Dental Students_sentence_1"
},
{
"operation": "finish",
"time": "501.393",
"value": "finish"
}
]
}
|
12256
|
Phil Bach was a footballer from Wales who played at full back with the England National Team in 1899.
|
{
"evidence": [
"Phil Bach_sentence_3"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nPhil Bach was a footballer from Wales who played at full back with the England National Team in 1899.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Phil Bach_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7551,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.052",
"value": "?search="
},
{
"operation": "search",
"time": "9.327",
"value": "Phil Bach"
},
{
"operation": "Now on",
"time": "9.87",
"value": "Phil Bach"
},
{
"operation": "Highlighting",
"time": "14.438",
"value": "Phil Bach_sentence_0"
},
{
"operation": "Page search",
"time": "35.451",
"value": "wales"
},
{
"operation": "Highlighting",
"time": "80.114",
"value": "Phil Bach_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "84.247",
"value": "Phil Bach_sentence_0"
},
{
"operation": "finish",
"time": "109.214",
"value": "finish"
}
]
}
|
88603
|
Yevhen Baryshnikov is a defensive midfielder, a position on the field between the team's defenders and forwards that travels the greatest distance and has the most possession during a game.
|
{
"evidence": [
"Yevhen Baryshnikov_sentence_0",
"Midfielder_sentence_3",
"Midfielder_sentence_8",
"Midfielder_sentence_9"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nYevhen Baryshnikov is a defensive midfielder, a position on the field between the team's defenders and forwards that travels the greatest distance and has the most possession during a game.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Yevhen Baryshnikov_sentence_0",
"Midfielder_sentence_3",
"Midfielder_sentence_8",
"Midfielder_sentence_9"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7552,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.135",
"value": "?search="
},
{
"operation": "search",
"time": "7.929",
"value": "Yevhen Baryshnikov"
},
{
"operation": "Now on",
"time": "8.813",
"value": "Yevhen Baryshnikov"
},
{
"operation": "Highlighting",
"time": "17.672",
"value": "Yevhen Baryshnikov_sentence_0"
},
{
"operation": "hyperlink",
"time": "58.833",
"value": "defensive midfielder"
},
{
"operation": "Now on",
"time": "59.838",
"value": "Midfielder"
},
{
"operation": "Highlighting",
"time": "65.172",
"value": "Midfielder_sentence_3"
},
{
"operation": "Highlighting",
"time": "90.132",
"value": "Midfielder_sentence_8"
},
{
"operation": "Highlighting",
"time": "93.362",
"value": "Midfielder_sentence_9"
},
{
"operation": "finish",
"time": "100.572",
"value": "finish"
}
]
}
|
exfever_4587
|
John Goodman received Best Actor – Television Series Musical or Comedy that was first awarded under the title Best TV Star – Male to a South African-born American television personality.
|
{
"evidence": [
"John Goodman",
"Golden Globe Award for Best Actor – Television Series Musical or Comedy"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJohn Goodman received Best Actor – Television Series Musical or Comedy that was first awarded under the title Best TV Star – Male to a South African-born American television personality.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"John Goodman",
"Golden Globe Award for Best Actor – Television Series Musical or Comedy"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7553,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
50335
|
William Ecker lived in the Virginia suburbs of Washington D.C. and moved to Punta Gorda, Florida in 2001 and finished writing a book with Kenneth V. Jack.
|
{
"evidence": [
"William Ecker_sentence_9",
"William Ecker_sentence_10",
"William Ecker_sentence_11",
"William Ecker_sentence_12"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nWilliam Ecker lived in the Virginia suburbs of Washington D.C. and moved to Punta Gorda, Florida in 2001 and finished writing a book with Kenneth V. Jack.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"William Ecker_sentence_9",
"William Ecker_sentence_10",
"William Ecker_sentence_11",
"William Ecker_sentence_12"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7554,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.435",
"value": "?search="
},
{
"operation": "search",
"time": "23.685",
"value": "William Ecker"
},
{
"operation": "Now on",
"time": "24.632",
"value": "William Ecker"
},
{
"operation": "Highlighting",
"time": "37.229",
"value": "William Ecker_sentence_9"
},
{
"operation": "Highlighting",
"time": "46.974",
"value": "William Ecker_sentence_10"
},
{
"operation": "Highlighting",
"time": "55.443",
"value": "William Ecker_sentence_11"
},
{
"operation": "search",
"time": "88.551",
"value": "Blue Moon Over Cuba"
},
{
"operation": "Now on",
"time": "91.225",
"value": "undefined"
},
{
"operation": "search",
"time": "111.519",
"value": "Kenneth V. Jack"
},
{
"operation": "Now on",
"time": "113.465",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "151.446",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "151.515",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "152.786",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "152.801",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "152.859",
"value": "William Ecker"
},
{
"operation": "Highlighting",
"time": "160.871",
"value": "William Ecker_sentence_12"
},
{
"operation": "finish",
"time": "166.697",
"value": "finish"
}
]
}
|
exfever_7111
|
The Lion King is a 1995 American animated drama film whose film genre is where the songs of the characters are interwoven into the narrative.
|
{
"evidence": [
"The Lion King",
"Musical film"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Lion King is a 1995 American animated drama film whose film genre is where the songs of the characters are interwoven into the narrative.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Lion King",
"Musical film"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7555,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3276
|
Cary Grant initially appeared in crime films or dramas such as an American pre-Code drama film starring an English stage, screen, and radio actor who starred in many popular well-regarded Hollywood films.
|
{
"evidence": [
"Cary Grant",
"Blonde Venus",
"Herbert Marshall"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCary Grant initially appeared in crime films or dramas such as an American pre-Code drama film starring an English stage, screen, and radio actor who starred in many popular well-regarded Hollywood films.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Cary Grant",
"Blonde Venus",
"Herbert Marshall"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7556,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
51498
|
Ulrome is a village in the East Riding of Yorkshire, situated approximately 6 miles (9.7 km) north of the town of Hornsea which is a sovereign state of Europe.
|
{
"evidence": [
"Ulrome_sentence_0",
"Ulrome_sentence_1",
"Hornsea_sentence_2"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nUlrome is a village in the East Riding of Yorkshire, situated approximately 6 miles (9.7 km) north of the town of Hornsea which is a sovereign state of Europe.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Ulrome_sentence_0",
"Ulrome_sentence_1",
"Hornsea_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7557,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.105",
"value": "?search="
},
{
"operation": "Now on",
"time": "8.873",
"value": "?search="
},
{
"operation": "search",
"time": "18.185",
"value": "Ulrome"
},
{
"operation": "Now on",
"time": "18.993",
"value": "Ulrome"
},
{
"operation": "Highlighting",
"time": "25.073",
"value": "Ulrome_sentence_0"
},
{
"operation": "Highlighting",
"time": "33.314",
"value": "Ulrome_sentence_1"
},
{
"operation": "hyperlink",
"time": "36.607",
"value": "Hornsea"
},
{
"operation": "Now on",
"time": "38.278",
"value": "Hornsea"
},
{
"operation": "Highlighting",
"time": "57.575",
"value": "Hornsea_sentence_2"
},
{
"operation": "finish",
"time": "334.748",
"value": "finish"
}
]
}
|
43473
|
Lough Sheelin lies on the early course of the River Inny, a major tributary of the Shannon; its inflow is sometimes known as the Upper Inny or the Ross River, while the outflow is known as the Lower Inny.
|
{
"evidence": [
"Lough Sheelin_sentence_1",
"River Inny (Leinster)_sentence_0",
"Lough Sheelin_sentence_5"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nLough Sheelin lies on the early course of the River Inny, a major tributary of the Shannon; its inflow is sometimes known as the Upper Inny or the Ross River, while the outflow is known as the Lower Inny.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Lough Sheelin_sentence_1",
"River Inny (Leinster)_sentence_0",
"Lough Sheelin_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7558,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.441",
"value": "?search="
},
{
"operation": "search",
"time": "6.889",
"value": "Lough Sheelin"
},
{
"operation": "Now on",
"time": "7.906",
"value": "Lough Sheelin"
},
{
"operation": "Highlighting",
"time": "20.005",
"value": "Lough Sheelin_sentence_1"
},
{
"operation": "hyperlink",
"time": "26.044",
"value": "River Inny"
},
{
"operation": "Now on",
"time": "26.94",
"value": "River Inny (Leinster)"
},
{
"operation": "Highlighting",
"time": "45.177",
"value": "River Inny (Leinster)_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "47.244",
"value": "River Inny (Leinster)_sentence_0"
},
{
"operation": "hyperlink",
"time": "47.805",
"value": "River Shannon"
},
{
"operation": "Now on",
"time": "49.221",
"value": "River Shannon"
},
{
"operation": "back-button-clicked",
"time": "55.47",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "55.599",
"value": "River Inny (Leinster)"
},
{
"operation": "Highlighting",
"time": "81.73",
"value": "River Inny (Leinster)_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "95.161",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "95.169",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "95.288",
"value": "Lough Sheelin"
},
{
"operation": "Highlighting",
"time": "296.4",
"value": "Lough Sheelin_sentence_5"
},
{
"operation": "finish",
"time": "316.802",
"value": "finish"
}
]
}
|
59567
|
Cities XL 2012 (released on October 20,2011) is the third installment in the Cities XL franchise and focuses on a single-player mode.
|
{
"evidence": [
"Cities XL 2012_sentence_1",
"Cities XL 2012_sentence_2"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCities XL 2012 (released on October 20,2011) is the third installment in the Cities XL franchise and focuses on a single-player mode.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Cities XL 2012_sentence_1",
"Cities XL 2012_sentence_2"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7559,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "3.648",
"value": "?search="
},
{
"operation": "search",
"time": "17.351",
"value": "Cities XL 2012"
},
{
"operation": "Now on",
"time": "22.345",
"value": "Cities XL 2012"
},
{
"operation": "Highlighting",
"time": "52.897",
"value": "Cities XL 2012_sentence_1"
},
{
"operation": "Highlighting",
"time": "64.177",
"value": "Cities XL 2012_sentence_2"
},
{
"operation": "Page search",
"time": "78.621",
"value": "install"
},
{
"operation": "Page search",
"time": "78.971",
"value": "install"
},
{
"operation": "Page search",
"time": "79.163",
"value": "install"
},
{
"operation": "Page search",
"time": "79.315",
"value": "install"
},
{
"operation": "finish",
"time": "122.064",
"value": "finish"
}
]
}
|
90524
|
In road and rail transport, and water and waste management networks, civil engineers design the infrastructure (but urban planners could develop the general layout of roads, sidewalks and public places); which they also do for electrical power and lighting networks and telecommunications, computing and monitoring networks.
|
{
"evidence": [
"Infrastructure and economics_sentence_33"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nIn road and rail transport, and water and waste management networks, civil engineers design the infrastructure (but urban planners could develop the general layout of roads, sidewalks and public places); which they also do for electrical power and lighting networks and telecommunications, computing and monitoring networks.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Infrastructure and economics_sentence_33"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7560,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.444",
"value": "?search="
},
{
"operation": "search",
"time": "209.325",
"value": "Municipal or urban engineering"
},
{
"operation": "Now on",
"time": "210.281",
"value": "Municipal or urban engineering"
},
{
"operation": "search",
"time": "815.146",
"value": "Infrastructure and economics"
},
{
"operation": "Now on",
"time": "816.094",
"value": "Infrastructure and economics"
},
{
"operation": "Highlighting",
"time": "819.252",
"value": "Infrastructure and economics_sentence_33"
},
{
"operation": "finish",
"time": "863.662",
"value": "finish"
}
]
}
|
6469
|
Clare Short was a Shadow Minister for Women, Shadow Transport Secretary, Opposition Spokesperson for Overseas Development, and a member of Labour's National Executive Committee (the governing body of the Conservative Party).
|
{
"evidence": [
"National Executive Committee of the Labour Party_sentence_0"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nClare Short was a Shadow Minister for Women, Shadow Transport Secretary, Opposition Spokesperson for Overseas Development, and a member of Labour's National Executive Committee (the governing body of the Conservative Party).\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"National Executive Committee of the Labour Party_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7561,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.969",
"value": "?search="
},
{
"operation": "search",
"time": "11.08",
"value": "Clare Short"
},
{
"operation": "Now on",
"time": "11.988",
"value": "Clare Short"
},
{
"operation": "Highlighting",
"time": "69.358",
"value": "Clare Short_sentence_29"
},
{
"operation": "Highlighting",
"time": "69.365",
"value": "Clare Short_title"
},
{
"operation": "Highlighting",
"time": "69.369",
"value": "Clare Short_section_0"
},
{
"operation": "Highlighting",
"time": "69.373",
"value": "Clare Short_section_2"
},
{
"operation": "Highlighting",
"time": "77.822",
"value": "Clare Short_sentence_30"
},
{
"operation": "hyperlink",
"time": "82.277",
"value": "National Executive Committee"
},
{
"operation": "Now on",
"time": "83.093",
"value": "National Executive Committee of the Labour Party"
},
{
"operation": "Highlighting",
"time": "87.726",
"value": "National Executive Committee of the Labour Party_sentence_0"
},
{
"operation": "finish",
"time": "116.819",
"value": "finish"
}
]
}
|
52617
|
Machilipatnam is a city in the Krishna district of the Andhra Pradesh state in India.
|
{
"evidence": [
"Machilipatnam_sentence_0"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMachilipatnam is a city in the Krishna district of the Andhra Pradesh state in India.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Machilipatnam_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7562,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.895",
"value": "?search="
},
{
"operation": "Now on",
"time": "9.608",
"value": "?search="
},
{
"operation": "search",
"time": "132.11",
"value": "Machilipatnam"
},
{
"operation": "Now on",
"time": "133.268",
"value": "Machilipatnam"
},
{
"operation": "Highlighting",
"time": "189.461",
"value": "Machilipatnam_sentence_0"
},
{
"operation": "finish",
"time": "217.106",
"value": "finish"
}
]
}
|
87485
|
Marc Salyers started his professional years playing for Legadue and later on became a top scorer on the 2003–04 TBL season.
|
{
"evidence": [
"Marc Salyers_sentence_4",
"Marc Salyers_sentence_6"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nMarc Salyers started his professional years playing for Legadue and later on became a top scorer on the 2003–04 TBL season.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Marc Salyers_sentence_4",
"Marc Salyers_sentence_6"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7563,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.35",
"value": "?search="
},
{
"operation": "search",
"time": "9.798",
"value": "Marc Salyers"
},
{
"operation": "Now on",
"time": "10.766",
"value": "Marc Salyers"
},
{
"operation": "Highlighting",
"time": "37.159",
"value": "Marc Salyers_sentence_4"
},
{
"operation": "Highlighting",
"time": "45.899",
"value": "Marc Salyers_sentence_6"
},
{
"operation": "finish",
"time": "54.787",
"value": "finish"
}
]
}
|
50448
|
Gender mainstreaming in mine action started way back the mid 1970s and later on the concept of gender mainstreaming appeared in the mid 1980s.
|
{
"evidence": [
"Gender mainstreaming in mine action_sentence_6"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nGender mainstreaming in mine action started way back the mid 1970s and later on the concept of gender mainstreaming appeared in the mid 1980s.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Gender mainstreaming in mine action_sentence_6"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7564,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "16.532",
"value": "?search="
},
{
"operation": "search",
"time": "316.797",
"value": "Gender mainstreaming in mine action"
},
{
"operation": "Now on",
"time": "319.519",
"value": "Gender mainstreaming in mine action"
},
{
"operation": "Highlighting",
"time": "586.586",
"value": "Gender mainstreaming in mine action_sentence_6"
},
{
"operation": "Highlighting",
"time": "587.328",
"value": "Gender mainstreaming in mine action_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "768.258",
"value": "Gender mainstreaming in mine action_sentence_7"
},
{
"operation": "finish",
"time": "831.031",
"value": "finish"
}
]
}
|
exfever_3678
|
Tom Felton is an English actor best known for playing Draco Malfoy in the film adaptations of the series of seven fantasy novels written by British author J.R.R. Tolkien.
|
{
"evidence": [
"Tom Felton",
"Harry Potter"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nTom Felton is an English actor best known for playing Draco Malfoy in the film adaptations of the series of seven fantasy novels written by British author J.R.R. Tolkien.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tom Felton",
"Harry Potter"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7565,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
86017
|
The 1964 Cambridge City Council election, which took place on 7 May 1964, was the first to the newly formed London Borough of Westminster and had a total of 155 candidates for the 60 seats being contested across 31 wards.
|
{
"evidence": [
"City of Westminster_sentence_7"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe 1964 Cambridge City Council election, which took place on 7 May 1964, was the first to the newly formed London Borough of Westminster and had a total of 155 candidates for the 60 seats being contested across 31 wards.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"City of Westminster_sentence_7"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7566,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.893",
"value": "?search="
},
{
"operation": "search",
"time": "9.53",
"value": "1964 Cambridge City Council election"
},
{
"operation": "Now on",
"time": "11.531",
"value": "undefined"
},
{
"operation": "search",
"time": "62.154",
"value": "Cambridge City Council elections"
},
{
"operation": "Now on",
"time": "63.204",
"value": "Cambridge City Council elections"
},
{
"operation": "hyperlink",
"time": "89.947",
"value": "Cambridge City Council"
},
{
"operation": "Now on",
"time": "90.519",
"value": "Cambridge City Council"
},
{
"operation": "Page search",
"time": "107.32",
"value": "1964"
},
{
"operation": "search",
"time": "132.504",
"value": "London Borough of Westminster"
},
{
"operation": "Now on",
"time": "134.192",
"value": "City of Westminster"
},
{
"operation": "Highlighting",
"time": "213.021",
"value": "City of Westminster_sentence_7"
},
{
"operation": "finish",
"time": "259.982",
"value": "finish"
}
]
}
|
exfever_6931
|
The National Security Agency is the first national-level security agency in the United States.
|
{
"evidence": [
"National Security Agency",
"Intelligence agency"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe National Security Agency is the first national-level security agency in the United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"National Security Agency",
"Intelligence agency"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7567,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5588
|
A spread is a food that is a spreadable condiment, such as a dairy product made from the fat and protein components of churned cream.
|
{
"evidence": [
"Spread (food)",
"Butter"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nA spread is a food that is a spreadable condiment, such as a dairy product made from the fat and protein components of churned cream.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Spread (food)",
"Butter"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7568,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
19407
|
It is privately owned at some point in the Oswego Recreational Trail, which hikers need to follow the right-of-way rule, while some of it has not been resurfaced yet, except in a few eroded spots, a few miles to Cleveland and New York.
|
{
"evidence": [
"Oswego Recreational Trail_sentence_10",
"Oswego Recreational Trail_sentence_11",
"Oswego Recreational Trail_sentence_9"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nIt is privately owned at some point in the Oswego Recreational Trail, which hikers need to follow the right-of-way rule, while some of it has not been resurfaced yet, except in a few eroded spots, a few miles to Cleveland and New York.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Oswego Recreational Trail_sentence_10",
"Oswego Recreational Trail_sentence_11",
"Oswego Recreational Trail_sentence_9"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7569,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.609",
"value": "?search="
},
{
"operation": "search",
"time": "244.867",
"value": "Oswego Recreational Trail"
},
{
"operation": "Now on",
"time": "245.806",
"value": "Oswego Recreational Trail"
},
{
"operation": "Highlighting",
"time": "256.379",
"value": "Oswego Recreational Trail_sentence_10"
},
{
"operation": "Highlighting",
"time": "265.705",
"value": "Oswego Recreational Trail_sentence_4"
},
{
"operation": "Highlighting",
"time": "278.088",
"value": "Oswego Recreational Trail_sentence_11"
},
{
"operation": "Highlighting",
"time": "285.314",
"value": "Oswego Recreational Trail_sentence_9"
},
{
"operation": "Highlighting deleted",
"time": "297.16",
"value": "Oswego Recreational Trail_sentence_4"
},
{
"operation": "finish",
"time": "315.334",
"value": "finish"
}
]
}
|
15797
|
RELB, sequestered by the NF-κB precursor protein p100 in the cytoplasm, appears in the nucleus with p70 and activates the expression homeostatic lymphokines.
|
{
"evidence": [
"RELB_sentence_4",
"RELB_sentence_5"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nRELB, sequestered by the NF-κB precursor protein p100 in the cytoplasm, appears in the nucleus with p70 and activates the expression homeostatic lymphokines.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"RELB_sentence_4",
"RELB_sentence_5"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7570,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.049",
"value": "?search="
},
{
"operation": "search",
"time": "17.622",
"value": "RELB"
},
{
"operation": "Now on",
"time": "18.291",
"value": "RELB"
},
{
"operation": "Page search",
"time": "29.813",
"value": "NF-κB"
},
{
"operation": "Highlighting",
"time": "36.33",
"value": "RELB_sentence_2"
},
{
"operation": "Highlighting",
"time": "48.891",
"value": "RELB_sentence_4"
},
{
"operation": "Highlighting",
"time": "87.702",
"value": "RELB_sentence_5"
},
{
"operation": "Page search",
"time": "96.447",
"value": "p70"
},
{
"operation": "Highlighting deleted",
"time": "108.525",
"value": "RELB_sentence_4"
},
{
"operation": "Highlighting",
"time": "109.552",
"value": "RELB_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "113.791",
"value": "RELB_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "120.443",
"value": "RELB_sentence_2"
},
{
"operation": "Highlighting",
"time": "129.758",
"value": "RELB_sentence_5"
},
{
"operation": "finish",
"time": "144.679",
"value": "finish"
}
]
}
|
68990
|
James Colebrook Patterson had a long career in politics including Sixth Lieutenant Governor of Manitoba and Member of the Legislative assembly of Ontario; he had no interests outside of politics.
|
{
"evidence": [
"James Colebrooke Patterson_sentence_4"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJames Colebrook Patterson had a long career in politics including Sixth Lieutenant Governor of Manitoba and Member of the Legislative assembly of Ontario; he had no interests outside of politics.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"James Colebrooke Patterson_sentence_4"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7571,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Highlighting",
"time": "0.003",
"value": "Chi2 Orionis_sentence_4"
},
{
"operation": "Highlighting",
"time": "0.003",
"value": "Chi2 Orionis_title"
},
{
"operation": "Now on",
"time": "0.005",
"value": "?search="
},
{
"operation": "Now on",
"time": "0.023",
"value": "?search="
},
{
"operation": "search",
"time": "0.039",
"value": "james colebrook patterson"
},
{
"operation": "Now on",
"time": "0.042",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "0.061",
"value": "james colebrooke patterson"
},
{
"operation": "Now on",
"time": "0.062",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "0.073",
"value": "Malcolm Colin Cameron"
},
{
"operation": "Now on",
"time": "0.075",
"value": "Malcolm Colin Cameron"
},
{
"operation": "hyperlink",
"time": "0.087",
"value": "James Colebrooke Patterson"
},
{
"operation": "Now on",
"time": "0.088",
"value": "James Colebrooke Patterson"
},
{
"operation": "Highlighting",
"time": "0.122",
"value": "James Colebrooke Patterson_sentence_1"
},
{
"operation": "Highlighting",
"time": "0.135",
"value": "James Colebrooke Patterson_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "0.156",
"value": "James Colebrooke Patterson_sentence_1"
},
{
"operation": "finish",
"time": "0.168",
"value": "finish"
},
{
"operation": "Now on",
"time": "1621887462.699",
"value": "Chi2 Orionis"
},
{
"operation": "finish",
"time": "1621887469.835",
"value": "finish"
}
]
}
|
exfever_8818
|
Andy Dwyer is a fictional character that eventually marries April Ludgate, portrayed by an American actress who starred in the NBC genre of comedy centered on a fixed set of characters.
|
{
"evidence": [
"Andy Dwyer",
"Aubrey Plaza",
"Sitcom"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAndy Dwyer is a fictional character that eventually marries April Ludgate, portrayed by an American actress who starred in the NBC genre of comedy centered on a fixed set of characters.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Andy Dwyer",
"Aubrey Plaza",
"Sitcom"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7572,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6443
|
Crystal Palace F.C. is a club of a team sport that is played in accordance with a set of codified rules that have been maintained by the IFAB.
|
{
"evidence": [
"Crystal Palace F.C.",
"Association football",
"Laws of the Game (association football)"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCrystal Palace F.C. is a club of a team sport that is played in accordance with a set of codified rules that have been maintained by the IFAB.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Crystal Palace F.C.",
"Association football",
"Laws of the Game (association football)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7573,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
2381
|
Finch Field, located in Thomasville, a city in Davidson County, North Carolina, United States, was demolished in 2000.
|
{
"evidence": [
"Finch Field_sentence_2"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nFinch Field, located in Thomasville, a city in Davidson County, North Carolina, United States, was demolished in 2000.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Finch Field_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7574,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "17.683",
"value": "Finch Field"
},
{
"operation": "Highlighting",
"time": "21.612",
"value": "Finch Field_sentence_0"
},
{
"operation": "hyperlink",
"time": "24.858",
"value": "Thomasville"
},
{
"operation": "Highlighting",
"time": "27.119",
"value": "Thomasville, North Carolina_sentence_0"
},
{
"operation": "Highlighting",
"time": "27.185",
"value": "Thomasville, North Carolina_title"
},
{
"operation": "back-button-clicked",
"time": "29.891",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "30.001",
"value": "back-button-clicked"
},
{
"operation": "search",
"time": "38.242",
"value": "Finch Field"
},
{
"operation": "Highlighting",
"time": "45.667",
"value": "Finch Field_sentence_2"
},
{
"operation": "finish",
"time": "83.741",
"value": "finish"
}
]
}
|
exfever_5687
|
The Avenger is a 2012 American superhero film distributed by the American film distribution studio in the Walt Disney Company's Disney Media and Entertainment Distribution division.
|
{
"evidence": [
"The Avengers (2012 film)",
"Walt Disney Studios Motion Pictures"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Avenger is a 2012 American superhero film distributed by the American film distribution studio in the Walt Disney Company's Disney Media and Entertainment Distribution division.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Avengers (2012 film)",
"Walt Disney Studios Motion Pictures"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7575,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
8351
|
Wanderers were the reigning cup holders in both 1872 and 1873, but Oxford University had won in 1874.
|
{
"evidence": [
"1877 FA Cup Final_sentence_10",
"1877 FA Cup Final_sentence_11"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nWanderers were the reigning cup holders in both 1872 and 1873, but Oxford University had won in 1874.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"1877 FA Cup Final_sentence_10",
"1877 FA Cup Final_sentence_11"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7576,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.76",
"value": "?search="
},
{
"operation": "search",
"time": "40.868",
"value": "1877 FA Cup Final"
},
{
"operation": "Now on",
"time": "41.78",
"value": "1877 FA Cup Final"
},
{
"operation": "Highlighting",
"time": "47.538",
"value": "1877 FA Cup Final_sentence_10"
},
{
"operation": "Highlighting",
"time": "50.733",
"value": "1877 FA Cup Final_sentence_11"
},
{
"operation": "finish",
"time": "61.941",
"value": "finish"
}
]
}
|
92250
|
CALCRL associated with RAMP1, it is a heterodimer protein and its sequence reveals multiple hydrophobic and hydrophilic regions.
|
{
"evidence": [
"CALCRL_sentence_9",
"CALCRL_sentence_11",
"CALCRL_sentence_12"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nCALCRL associated with RAMP1, it is a heterodimer protein and its sequence reveals multiple hydrophobic and hydrophilic regions.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"CALCRL_sentence_9",
"CALCRL_sentence_11",
"CALCRL_sentence_12"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7577,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.368",
"value": "?search="
},
{
"operation": "search",
"time": "9.635",
"value": "CALCRL"
},
{
"operation": "Now on",
"time": "10.958",
"value": "CALCRL"
},
{
"operation": "Highlighting",
"time": "22.991",
"value": "CALCRL_sentence_2"
},
{
"operation": "Highlighting",
"time": "22.992",
"value": "CALCRL_section_0"
},
{
"operation": "Highlighting deleted",
"time": "34.24",
"value": "CALCRL_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "34.245",
"value": "CALCRL_section_0"
},
{
"operation": "Page search",
"time": "46.865",
"value": "heterodimer"
},
{
"operation": "Highlighting",
"time": "54.176",
"value": "CALCRL_sentence_9"
},
{
"operation": "Highlighting",
"time": "59.854",
"value": "CALCRL_sentence_11"
},
{
"operation": "Highlighting",
"time": "62.61",
"value": "CALCRL_sentence_12"
},
{
"operation": "finish",
"time": "71.172",
"value": "finish"
}
]
}
|
exfever_951
|
The Pennsylvania Chronicle was an American colonial newspaper founded by William Goddard and his business partner, an American attorney who become one of the most prominent traitors to the American cause during the early part of the Revolutionary War.
|
{
"evidence": [
"Pennsylvania Chronicle",
"Joseph Galloway"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Pennsylvania Chronicle was an American colonial newspaper founded by William Goddard and his business partner, an American attorney who become one of the most prominent traitors to the American cause during the early part of the Revolutionary War.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Pennsylvania Chronicle",
"Joseph Galloway"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7578,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
26854
|
After losing the first five games of the 1931–32 regular season, the Toronto Maple Leafs, whose colours are navy blue and white, fired head coach Art Duncan and replaced him with Dick Irvin, who had an 8–3–2 record in his first month behind the bench.
|
{
"evidence": [
"1931–32 Toronto Maple Leafs season_sentence_5",
"1931–32 Toronto Maple Leafs season_sentence_6",
"1931–32 Toronto Maple Leafs season_sentence_8"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAfter losing the first five games of the 1931–32 regular season, the Toronto Maple Leafs, whose colours are navy blue and white, fired head coach Art Duncan and replaced him with Dick Irvin, who had an 8–3–2 record in his first month behind the bench.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"1931–32 Toronto Maple Leafs season_sentence_5",
"1931–32 Toronto Maple Leafs season_sentence_6",
"1931–32 Toronto Maple Leafs season_sentence_8"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7579,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.912",
"value": "?search="
},
{
"operation": "search",
"time": "15.203",
"value": "Toronto Maple Leafs"
},
{
"operation": "Now on",
"time": "21.341",
"value": "Toronto Maple Leafs"
},
{
"operation": "hyperlink",
"time": "46.541",
"value": "1.2.1 Opening of Maple Leaf Gardens (1930s) Toronto Maple Leafs_section_3"
},
{
"operation": "Page search",
"time": "98.944",
"value": "art"
},
{
"operation": "Page search",
"time": "117.407",
"value": "art duncan"
},
{
"operation": "Page search",
"time": "143.978",
"value": "fir"
},
{
"operation": "Page search",
"time": "144.601",
"value": "fire"
},
{
"operation": "hyperlink",
"time": "238.212",
"value": "Dick Irvin"
},
{
"operation": "Now on",
"time": "240.889",
"value": "Dick Irvin"
},
{
"operation": "Page search",
"time": "289.738",
"value": "coach"
},
{
"operation": "hyperlink",
"time": "337.046",
"value": "Conn Smythe"
},
{
"operation": "Now on",
"time": "339.981",
"value": "Conn Smythe"
},
{
"operation": "search",
"time": "353.561",
"value": "art duncan"
},
{
"operation": "Now on",
"time": "354.833",
"value": "Art Duncan"
},
{
"operation": "Highlighting",
"time": "419.021",
"value": "Art Duncan_sentence_29"
},
{
"operation": "Highlighting",
"time": "419.026",
"value": "Art Duncan_title"
},
{
"operation": "Highlighting",
"time": "419.028",
"value": "Art Duncan_section_2"
},
{
"operation": "Highlighting deleted",
"time": "422.608",
"value": "Art Duncan_sentence_29"
},
{
"operation": "Highlighting deleted",
"time": "422.615",
"value": "Art Duncan_title"
},
{
"operation": "Highlighting deleted",
"time": "422.619",
"value": "Art Duncan_section_2"
},
{
"operation": "Highlighting",
"time": "424.217",
"value": "Art Duncan_sentence_29"
},
{
"operation": "Highlighting",
"time": "424.223",
"value": "Art Duncan_title"
},
{
"operation": "Highlighting",
"time": "424.224",
"value": "Art Duncan_section_2"
},
{
"operation": "hyperlink",
"time": "426.664",
"value": "Toronto Maple Leafs"
},
{
"operation": "Now on",
"time": "427.957",
"value": "Toronto Maple Leafs"
},
{
"operation": "Page search",
"time": "443.387",
"value": "1931"
},
{
"operation": "search",
"time": "561.627",
"value": "1931-32 Toronto Maple Leafs"
},
{
"operation": "Now on",
"time": "565.364",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "567.566",
"value": "1931–32 Toronto Maple Leafs season"
},
{
"operation": "Now on",
"time": "572.255",
"value": "1931–32 Toronto Maple Leafs season"
},
{
"operation": "Highlighting",
"time": "582.787",
"value": "1931–32 Toronto Maple Leafs season_sentence_5"
},
{
"operation": "Highlighting",
"time": "585.334",
"value": "1931–32 Toronto Maple Leafs season_sentence_6"
},
{
"operation": "Highlighting",
"time": "606.47",
"value": "1931–32 Toronto Maple Leafs season_sentence_8"
},
{
"operation": "finish",
"time": "634.752",
"value": "finish"
}
]
}
|
72515
|
The Brahman Brothers (identical twins and professional wrestlers) placed third in the UWA World Tag Team Championship.
|
{
"evidence": [
"The Brahman Brothers_sentence_2",
"The Brahman Brothers_sentence_0"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Brahman Brothers (identical twins and professional wrestlers) placed third in the UWA World Tag Team Championship.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"The Brahman Brothers_sentence_2",
"The Brahman Brothers_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 7580,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.188",
"value": "?search="
},
{
"operation": "search",
"time": "10.79",
"value": "Brahman Brothers"
},
{
"operation": "Now on",
"time": "11.659",
"value": "The Brahman Brothers"
},
{
"operation": "Highlighting",
"time": "15.315",
"value": "The Brahman Brothers_sentence_0"
},
{
"operation": "Highlighting",
"time": "39.084",
"value": "The Brahman Brothers_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "40.506",
"value": "The Brahman Brothers_sentence_0"
},
{
"operation": "Page search",
"time": "56.474",
"value": "third"
},
{
"operation": "Highlighting",
"time": "67.581",
"value": "The Brahman Brothers_sentence_0"
},
{
"operation": "finish",
"time": "329.398",
"value": "finish"
}
]
}
|
exfever_476
|
Jeffrey Nero Hardy is an American professional wrestler widely known for his time with World Wrestling Entertainment, Inc., an American professional boxing promotion.
|
{
"evidence": [
"Jeff Hardy",
"WWE"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJeffrey Nero Hardy is an American professional wrestler widely known for his time with World Wrestling Entertainment, Inc., an American professional boxing promotion.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Jeff Hardy",
"WWE"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 7581,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_7978
|
The Punisher is an American television series derived from the already existing work of the first Marvel Netflix series, Daredevil.
|
{
"evidence": [
"The Punisher (TV series)",
"Spin-off (media)"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Punisher is an American television series derived from the already existing work of the first Marvel Netflix series, Daredevil.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Punisher (TV series)",
"Spin-off (media)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7582,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
3101
|
Hassanat Akinwande started her career in the 1980s and eventually starred in over 50 videos but was arrested in 2006 for possession of cocaine, a strong stimulant most frequently used as a recreational drug.
|
{
"evidence": [
"Hassanat Akinwande_sentence_2",
"Hassanat Akinwande_sentence_3",
"Hassanat Akinwande_sentence_4",
"Cocaine_sentence_1"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHassanat Akinwande started her career in the 1980s and eventually starred in over 50 videos but was arrested in 2006 for possession of cocaine, a strong stimulant most frequently used as a recreational drug.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Hassanat Akinwande_sentence_2",
"Hassanat Akinwande_sentence_3",
"Hassanat Akinwande_sentence_4",
"Cocaine_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 7583,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "12.274",
"value": "Hassanat Akinwande"
},
{
"operation": "Highlighting",
"time": "15.867",
"value": "Hassanat Akinwande_sentence_2"
},
{
"operation": "Highlighting",
"time": "19.331",
"value": "Hassanat Akinwande_sentence_3"
},
{
"operation": "Highlighting",
"time": "22.504",
"value": "Hassanat Akinwande_sentence_4"
},
{
"operation": "search",
"time": "32.237",
"value": "cocaine"
},
{
"operation": "Highlighting",
"time": "37.432",
"value": "Cocaine_sentence_1"
},
{
"operation": "finish",
"time": "53.553",
"value": "finish"
}
]
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.