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_5650
|
Michelle LaVaughn Robinson Obama is an American attorney who graduated from a private research university in Cambridge, Massachusetts, founded in 1820.
|
{
"evidence": [
"Michelle Obama",
"Harvard Law School"
],
"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:\nMichelle LaVaughn Robinson Obama is an American attorney who graduated from a private research university in Cambridge, Massachusetts, founded in 1820.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Michelle Obama",
"Harvard Law School"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5784,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
73938
|
American players Herman Keiser and Byron Nelson never played any sport besides hockey.
|
{
"evidence": [
"Herman Keiser_sentence_0",
"Byron Nelson_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:\nAmerican players Herman Keiser and Byron Nelson never played any sport besides hockey.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Herman Keiser_sentence_0",
"Byron Nelson_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5785,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.573",
"value": "?search="
},
{
"operation": "search",
"time": "19.76",
"value": "Herman Keiser"
},
{
"operation": "Now on",
"time": "21.006",
"value": "Herman Keiser"
},
{
"operation": "Highlighting",
"time": "25.655",
"value": "Herman Keiser_sentence_0"
},
{
"operation": "search",
"time": "67.259",
"value": "Byron Nelson"
},
{
"operation": "Now on",
"time": "69.392",
"value": "Byron Nelson"
},
{
"operation": "Highlighting",
"time": "70.982",
"value": "Byron Nelson_sentence_0"
},
{
"operation": "finish",
"time": "86.999",
"value": "finish"
}
]
}
|
48284
|
Bull Lake Glacier(derived from the well-preserved moraines found in the vicinity of Bull Lake near the Wind River Mountains) is a 250 acres (100 ha) mountain glacier located in Fitzpatrick Wilderness of Shoshone National Forest, east of the Continental Divide in the American Rocky Mountains in Colorado.
|
{
"evidence": [
"Bull Lake Glacier_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:\nBull Lake Glacier(derived from the well-preserved moraines found in the vicinity of Bull Lake near the Wind River Mountains) is a 250 acres (100 ha) mountain glacier located in Fitzpatrick Wilderness of Shoshone National Forest, east of the Continental Divide in the American Rocky Mountains in Colorado.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Bull Lake Glacier_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5786,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.65",
"value": "?search="
},
{
"operation": "search",
"time": "10.463",
"value": "Bull Lake Glacier"
},
{
"operation": "Now on",
"time": "11.147",
"value": "Bull Lake Glacier"
},
{
"operation": "hyperlink",
"time": "43.498",
"value": "List of glaciers in the United States"
},
{
"operation": "Now on",
"time": "45.176",
"value": "List of glaciers in the United States"
},
{
"operation": "hyperlink",
"time": "86.569",
"value": "List of glaciers"
},
{
"operation": "Now on",
"time": "87.421",
"value": "List of glaciers"
},
{
"operation": "hyperlink",
"time": "103.242",
"value": "List of glaciers in the United States"
},
{
"operation": "Now on",
"time": "103.705",
"value": "List of glaciers in the United States"
},
{
"operation": "search",
"time": "122.709",
"value": "Bull Lake"
},
{
"operation": "Now on",
"time": "123.334",
"value": "Bull Lake"
},
{
"operation": "hyperlink",
"time": "126.786",
"value": "1 Glaciers Bull Lake_section_0"
},
{
"operation": "hyperlink",
"time": "130.86",
"value": "Bull Lake Glacier"
},
{
"operation": "Now on",
"time": "131.515",
"value": "Bull Lake Glacier"
},
{
"operation": "Highlighting",
"time": "146.723",
"value": "Bull Lake Glacier_cell_0_3_1"
},
{
"operation": "Highlighting",
"time": "146.73",
"value": "Bull Lake Glacier_header_cell_0_3_0"
},
{
"operation": "Highlighting",
"time": "146.73",
"value": "Bull Lake Glacier_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "154.515",
"value": "Bull Lake Glacier_sentence_1"
},
{
"operation": "Highlighting",
"time": "163.468",
"value": "Bull Lake Glacier_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "188.611",
"value": "Bull Lake Glacier_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "189.784",
"value": "Bull Lake Glacier_cell_0_3_1"
},
{
"operation": "Highlighting deleted",
"time": "189.787",
"value": "Bull Lake Glacier_header_cell_0_3_0"
},
{
"operation": "Highlighting deleted",
"time": "189.788",
"value": "Bull Lake Glacier_header_cell_0_0_0"
},
{
"operation": "finish",
"time": "202.883",
"value": "finish"
}
]
}
|
exfever_988
|
American Gods was the highest-rated television series in 2017.
|
{
"evidence": [
"American Gods (TV series)",
"American Gods"
],
"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:\nAmerican Gods was the highest-rated television series in 2017.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"American Gods (TV series)",
"American Gods"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5787,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
77167
|
Joacim Tuuri is a Finnish footballer who currently plays for Vaasan Palloseura.
|
{
"evidence": [
"Joacim Tuuri_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:\nJoacim Tuuri is a Finnish footballer who currently plays for Vaasan Palloseura.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Joacim Tuuri_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5788,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.192",
"value": "?search="
},
{
"operation": "search",
"time": "14.738",
"value": "Joacim Tuuri"
},
{
"operation": "Now on",
"time": "15.557",
"value": "Joacim Tuuri"
},
{
"operation": "Highlighting",
"time": "18.008",
"value": "Joacim Tuuri_sentence_0"
},
{
"operation": "finish",
"time": "213.595",
"value": "finish"
}
]
}
|
29077
|
Auguste Marie Fabre was a French industrialist with uptopian ideals and a devotion to the cooperative movement.
|
{
"evidence": [
"Auguste Marie Fabre_sentence_0",
"Auguste Marie Fabre_sentence_1",
"Auguste Marie Fabre_sentence_19"
],
"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:\nAuguste Marie Fabre was a French industrialist with uptopian ideals and a devotion to the cooperative movement.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Auguste Marie Fabre_sentence_0",
"Auguste Marie Fabre_sentence_1",
"Auguste Marie Fabre_sentence_19"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5789,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.823",
"value": "?search="
},
{
"operation": "search",
"time": "13.058",
"value": "Auguste Marie Fabre"
},
{
"operation": "Now on",
"time": "14.374",
"value": "Auguste Marie Fabre"
},
{
"operation": "Highlighting",
"time": "16.557",
"value": "Auguste Marie Fabre_sentence_0"
},
{
"operation": "Highlighting",
"time": "20.716",
"value": "Auguste Marie Fabre_sentence_1"
},
{
"operation": "Highlighting",
"time": "44.45",
"value": "Auguste Marie Fabre_sentence_19"
},
{
"operation": "finish",
"time": "80.302",
"value": "finish"
}
]
}
|
41913
|
Conway (born neither on March 16, 1930 nor on August 3, 2012) excelled in football, track and field, basketball, and baseball as a student at North Kansas City High School; then upon graduating, he was considered by many to be the best running back in football in the Kansas City.
|
{
"evidence": [
"Al Conway_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:\nConway (born neither on March 16, 1930 nor on August 3, 2012) excelled in football, track and field, basketball, and baseball as a student at North Kansas City High School; then upon graduating, he was considered by many to be the best running back in football in the Kansas City.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Al Conway_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5790,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.283",
"value": "?search="
},
{
"operation": "search",
"time": "20.452",
"value": "\"excelled in football, track and field, basketball, and baseball as a student \""
},
{
"operation": "Now on",
"time": "21.673",
"value": "undefined"
},
{
"operation": "search",
"time": "33.492",
"value": "\"he was considered by many to be the best running back in football\""
},
{
"operation": "Now on",
"time": "35.55",
"value": "undefined"
},
{
"operation": "search",
"time": "50.549",
"value": "\"student at North Kansas City High School\""
},
{
"operation": "Now on",
"time": "51.871",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "53.549",
"value": "Al Conway"
},
{
"operation": "Now on",
"time": "54.546",
"value": "Al Conway"
},
{
"operation": "Highlighting",
"time": "59.405",
"value": "Al Conway_sentence_0"
},
{
"operation": "finish",
"time": "89.017",
"value": "finish"
}
]
}
|
exfever_8700
|
Jesus was a first-century Jewish preacher and religious leader and believed to be the awaited messiah. Messiah is a saviour or liberator of a group of people which is a king or High Priest traditionally anointed with holy anointing oil according to the concepts that originated as an Abrahamic religious belief.
|
{
"evidence": [
"Jesus",
"Messiah",
"Messianism"
],
"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:\nJesus was a first-century Jewish preacher and religious leader and believed to be the awaited messiah. Messiah is a saviour or liberator of a group of people which is a king or High Priest traditionally anointed with holy anointing oil according to the concepts that originated as an Abrahamic religious belief.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Jesus",
"Messiah",
"Messianism"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5791,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_581
|
Stone Temple Pilots is a rock band with a former member and vocalist who was an Irish singer and songwriter and was also the lead vocalist of the bands Grey Daze, Dead by Sunrise, and Stone Temple Pilots.
|
{
"evidence": [
"Stone Temple Pilots",
"Chester Bennington"
],
"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:\nStone Temple Pilots is a rock band with a former member and vocalist who was an Irish singer and songwriter and was also the lead vocalist of the bands Grey Daze, Dead by Sunrise, and Stone Temple Pilots.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Stone Temple Pilots",
"Chester Bennington"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5792,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
43984
|
Instituto Mexicano de Ejecutivos de Finanzas ( In 1964 the IMEF research foundation awarded the first investigation contest, and In 2003 the IMEF joined Deloitte to spread the financial research applicable in Mexico and Latin America through this international open contest that has for objective to promote and encourage research on topics of interest to the macro-financial environment, as well as business and public finance)
|
{
"evidence": [
"Instituto Mexicano de Ejecutivos de Finanzas_sentence_4",
"Instituto Mexicano de Ejecutivos de Finanzas_sentence_15"
],
"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:\nInstituto Mexicano de Ejecutivos de Finanzas ( In 1964 the IMEF research foundation awarded the first investigation contest, and In 2003 the IMEF joined Deloitte to spread the financial research applicable in Mexico and Latin America through this international open contest that has for objective to promote and encourage research on topics of interest to the macro-financial environment, as well as business and public finance)\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Instituto Mexicano de Ejecutivos de Finanzas_sentence_4",
"Instituto Mexicano de Ejecutivos de Finanzas_sentence_15"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5793,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.075",
"value": "?search="
},
{
"operation": "search",
"time": "12.803",
"value": "Instituto Mexicano de Ejecutivos de Finanzas"
},
{
"operation": "Now on",
"time": "13.467",
"value": "Instituto Mexicano de Ejecutivos de Finanzas"
},
{
"operation": "Highlighting",
"time": "127.303",
"value": "Instituto Mexicano de Ejecutivos de Finanzas_sentence_4"
},
{
"operation": "Highlighting",
"time": "405.284",
"value": "Instituto Mexicano de Ejecutivos de Finanzas_sentence_15"
},
{
"operation": "finish",
"time": "536.86",
"value": "finish"
}
]
}
|
42003
|
The General Coordinating Committee of the Ministerial Conference on the Protection of Forests in Europe has five representatives from the signatory countries and serves as the financing body for the operation of its Liaison Unit Bratislava, currently based in Slovakia.
|
{
"evidence": [
"Ministerial Conference on the Protection of Forests in Europe_sentence_42"
],
"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 General Coordinating Committee of the Ministerial Conference on the Protection of Forests in Europe has five representatives from the signatory countries and serves as the financing body for the operation of its Liaison Unit Bratislava, currently based in Slovakia.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Ministerial Conference on the Protection of Forests in Europe_sentence_42"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5794,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.001",
"value": "?search="
},
{
"operation": "search",
"time": "0.009",
"value": "General Coordinating Committee of the Ministerial Conference on the Protection of Forests in Europe"
},
{
"operation": "Now on",
"time": "0.011",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "0.028",
"value": "Ministerial Conference on the Protection of Forests in Europe"
},
{
"operation": "Now on",
"time": "0.028",
"value": "Ministerial Conference on the Protection of Forests in Europe"
},
{
"operation": "Highlighting",
"time": "0.041",
"value": "Ministerial Conference on the Protection of Forests in Europe_sentence_42"
},
{
"operation": "finish",
"time": "0.311",
"value": "finish"
},
{
"operation": "Now on",
"time": "1620744052.252",
"value": "Ministerial Conference on the Protection of Forests in Europe"
},
{
"operation": "finish",
"time": "1620744054.779",
"value": "finish"
}
]
}
|
13808
|
Season of the Sparks received a rating of Orange from State, while The Irish Times (an Irish daily broadsheet newspaper and online digital publication launched on March 29, 1859) failed to give it a rating.
|
{
"evidence": [
"Season of the Sparks_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:\nSeason of the Sparks received a rating of Orange from State, while The Irish Times (an Irish daily broadsheet newspaper and online digital publication launched on March 29, 1859) failed to give it a rating.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Season of the Sparks_sentence_5"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5795,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.007",
"value": "?search="
},
{
"operation": "search",
"time": "19.778",
"value": "Season of the Sparks"
},
{
"operation": "Now on",
"time": "20.751",
"value": "Season of the Sparks"
},
{
"operation": "Page search",
"time": "72.773",
"value": "Orange from State"
},
{
"operation": "Highlighting",
"time": "89.444",
"value": "Season of the Sparks_cell_0_6_0"
},
{
"operation": "Highlighting",
"time": "89.461",
"value": "Season of the Sparks_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "89.462",
"value": "Season of the Sparks_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "90.173",
"value": "Season of the Sparks_cell_0_6_1"
},
{
"operation": "Highlighting",
"time": "90.19",
"value": "Season of the Sparks_header_cell_0_1_1"
},
{
"operation": "Highlighting",
"time": "97.515",
"value": "Season of the Sparks_cell_0_4_0"
},
{
"operation": "Highlighting",
"time": "98.164",
"value": "Season of the Sparks_cell_0_4_1"
},
{
"operation": "Highlighting",
"time": "107.511",
"value": "Season of the Sparks_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "109.891",
"value": "Season of the Sparks_cell_0_4_0"
},
{
"operation": "Highlighting deleted",
"time": "110.435",
"value": "Season of the Sparks_cell_0_4_1"
},
{
"operation": "Highlighting",
"time": "119.091",
"value": "Season of the Sparks_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "122.439",
"value": "Season of the Sparks_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "124.185",
"value": "Season of the Sparks_cell_0_6_1"
},
{
"operation": "Highlighting deleted",
"time": "124.197",
"value": "Season of the Sparks_header_cell_0_1_1"
},
{
"operation": "Highlighting deleted",
"time": "124.724",
"value": "Season of the Sparks_cell_0_6_0"
},
{
"operation": "Highlighting deleted",
"time": "124.739",
"value": "Season of the Sparks_header_cell_0_1_0"
},
{
"operation": "Highlighting deleted",
"time": "124.743",
"value": "Season of the Sparks_header_cell_0_0_0"
},
{
"operation": "hyperlink",
"time": "128.07",
"value": "The Irish Times"
},
{
"operation": "Now on",
"time": "129.364",
"value": "The Irish Times"
},
{
"operation": "back-button-clicked",
"time": "130.734",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "130.952",
"value": "Season of the Sparks"
},
{
"operation": "hyperlink",
"time": "132.94",
"value": "The Irish Times"
},
{
"operation": "Now on",
"time": "133.511",
"value": "The Irish Times"
},
{
"operation": "back-button-clicked",
"time": "135.21",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "135.41",
"value": "Season of the Sparks"
},
{
"operation": "finish",
"time": "145.939",
"value": "finish"
}
]
}
|
exfever_2024
|
Nicole Scherzinger studied musical theatre at a public research university in Fairborn, Ohio.
|
{
"evidence": [
"Nicole Scherzinger",
"Wright State University"
],
"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:\nNicole Scherzinger studied musical theatre at a public research university in Fairborn, Ohio.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Nicole Scherzinger",
"Wright State University"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5796,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_574
|
Judd Mann Apatow directed a 2015 American romantic comedy film with an ensemble cast that includes an American actor who co-starred in the drama The Stanford Prison Experiment.
|
{
"evidence": [
"Judd Apatow",
"Trainwreck (film)",
"Ezra Miller"
],
"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:\nJudd Mann Apatow directed a 2015 American romantic comedy film with an ensemble cast that includes an American actor who co-starred in the drama The Stanford Prison Experiment.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Judd Apatow",
"Trainwreck (film)",
"Ezra Miller"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5797,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
68608
|
Nic Endo was born in United States which has 50 states, he lived in Munich and joined ATR while they were on tour in 1997.
|
{
"evidence": [
"Nic Endo_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:\nNic Endo was born in United States which has 50 states, he lived in Munich and joined ATR while they were on tour in 1997.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Nic Endo_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5798,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.955",
"value": "?search="
},
{
"operation": "search",
"time": "9.668",
"value": "Nic Endo"
},
{
"operation": "Now on",
"time": "10.935",
"value": "Nic Endo"
},
{
"operation": "Highlighting",
"time": "34.452",
"value": "Nic Endo_cell_0_3_1"
},
{
"operation": "Highlighting",
"time": "34.457",
"value": "Nic Endo_header_cell_0_3_0"
},
{
"operation": "Highlighting",
"time": "34.458",
"value": "Nic Endo_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "34.458",
"value": "Nic Endo_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "54.99",
"value": "Nic Endo_sentence_3"
},
{
"operation": "Highlighting",
"time": "58.423",
"value": "Nic Endo_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "60.331",
"value": "Nic Endo_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "62.007",
"value": "Nic Endo_cell_0_3_1"
},
{
"operation": "Highlighting deleted",
"time": "62.011",
"value": "Nic Endo_header_cell_0_3_0"
},
{
"operation": "Highlighting deleted",
"time": "62.012",
"value": "Nic Endo_header_cell_0_1_0"
},
{
"operation": "Highlighting deleted",
"time": "62.014",
"value": "Nic Endo_header_cell_0_0_0"
},
{
"operation": "finish",
"time": "159.017",
"value": "finish"
}
]
}
|
76112
|
According to Lisan al-Arab, the term "muwalad" has a meaning close to "the adopted", and does not denote being of mixed-race but rather being neither of foreign-blood nor local culture.
|
{
"evidence": [
"Mulatto_sentence_21"
],
"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:\nAccording to Lisan al-Arab, the term \"muwalad\" has a meaning close to \"the adopted\", and does not denote being of mixed-race but rather being neither of foreign-blood nor local culture.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Mulatto_sentence_21"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5799,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.047",
"value": "?search="
},
{
"operation": "search",
"time": "10.436",
"value": "Lisan al-Arab"
},
{
"operation": "Now on",
"time": "11.979",
"value": "Lisan al-Arab"
},
{
"operation": "Page search",
"time": "18.419",
"value": "muwalad"
},
{
"operation": "search",
"time": "31.912",
"value": "Contains...muwalad"
},
{
"operation": "Now on",
"time": "32.913",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "36.979",
"value": "Mulatto"
},
{
"operation": "Now on",
"time": "37.834",
"value": "Mulatto"
},
{
"operation": "Page search",
"time": "40.597",
"value": "muwalad"
},
{
"operation": "Highlighting",
"time": "44.997",
"value": "Mulatto_sentence_19"
},
{
"operation": "Highlighting",
"time": "47.666",
"value": "Mulatto_sentence_20"
},
{
"operation": "Highlighting",
"time": "51.91",
"value": "Mulatto_sentence_21"
},
{
"operation": "Highlighting deleted",
"time": "58.466",
"value": "Mulatto_sentence_20"
},
{
"operation": "Highlighting deleted",
"time": "59.872",
"value": "Mulatto_sentence_19"
},
{
"operation": "finish",
"time": "78.452",
"value": "finish"
}
]
}
|
exfever_2763
|
Terrence Malick was also nominated for the Writers Guild of America Awards.
|
{
"evidence": [
"Sleeping Beauty (2011 film)",
"2011 Cannes Film Festival",
"Terrence Malick"
],
"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:\nTerrence Malick was also nominated for the Writers Guild of America Awards.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Sleeping Beauty (2011 film)",
"2011 Cannes Film Festival",
"Terrence Malick"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5800,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
75852
|
Uganda, with an IPC code of UGA, failed to send a delegation to the 1996 Summer Paralympics in Atlanta (the capital and most populous city of the U.S. state of Georgia).
|
{
"evidence": [
"Uganda at the 1996 Summer Paralympics_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:\nUganda, with an IPC code of UGA, failed to send a delegation to the 1996 Summer Paralympics in Atlanta (the capital and most populous city of the U.S. state of Georgia).\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Uganda at the 1996 Summer Paralympics_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5801,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.894",
"value": "?search="
},
{
"operation": "search",
"time": "6.478",
"value": "Uganda"
},
{
"operation": "Now on",
"time": "8.141",
"value": "Uganda"
},
{
"operation": "search",
"time": "56",
"value": "Uganda at the 1996 Summer Paralympics"
},
{
"operation": "Now on",
"time": "56.682",
"value": "Uganda at the 1996 Summer Paralympics"
},
{
"operation": "Highlighting",
"time": "68.355",
"value": "Uganda at the 1996 Summer Paralympics_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "69.148",
"value": "Uganda at the 1996 Summer Paralympics_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "69.581",
"value": "Uganda at the 1996 Summer Paralympics_cell_0_1_1"
},
{
"operation": "Highlighting",
"time": "69.59",
"value": "Uganda at the 1996 Summer Paralympics_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "69.592",
"value": "Uganda at the 1996 Summer Paralympics_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "76.168",
"value": "Uganda at the 1996 Summer Paralympics_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "83.67",
"value": "Uganda at the 1996 Summer Paralympics_cell_0_1_1"
},
{
"operation": "Highlighting deleted",
"time": "83.675",
"value": "Uganda at the 1996 Summer Paralympics_header_cell_0_1_0"
},
{
"operation": "Highlighting deleted",
"time": "83.677",
"value": "Uganda at the 1996 Summer Paralympics_header_cell_0_0_0"
},
{
"operation": "finish",
"time": "111.932",
"value": "finish"
}
]
}
|
56371
|
Dalton Moreira Neto (born February 5, 1990) is a Brazilian football player who currently plays for Luverdense.
|
{
"evidence": [
"Dalton Moreira Neto_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:\nDalton Moreira Neto (born February 5, 1990) is a Brazilian football player who currently plays for Luverdense.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Dalton Moreira Neto_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5802,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.727",
"value": "?search="
},
{
"operation": "search",
"time": "28.937",
"value": "Dalton Moreira Neto"
},
{
"operation": "Now on",
"time": "29.973",
"value": "Dalton Moreira Neto"
},
{
"operation": "Highlighting",
"time": "33.78",
"value": "Dalton Moreira Neto_sentence_0"
},
{
"operation": "finish",
"time": "46.342",
"value": "finish"
}
]
}
|
exfever_948
|
Kingdom Hearts III is a 2019 action role-playing game developed and published by Square Enix for a home video game console developed by Nintendo.
|
{
"evidence": [
"Kingdom Hearts III",
"Xbox One"
],
"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:\nKingdom Hearts III is a 2019 action role-playing game developed and published by Square Enix for a home video game console developed by Nintendo.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Kingdom Hearts III",
"Xbox One"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5803,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_823
|
Lost Highway is a 1997 neo-noir film, starring an American actor who also starred in a 1987 American space film directed by Mel Gibson.
|
{
"evidence": [
"Lost Highway (film)",
"Bill Pullman",
"Spaceballs"
],
"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:\nLost Highway is a 1997 neo-noir film, starring an American actor who also starred in a 1987 American space film directed by Mel Gibson.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Lost Highway (film)",
"Bill Pullman",
"Spaceballs"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5804,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
24294
|
An Indian company running Lexulous, formerly known as Scrabulous was threatened to be shut down before Facebook was asked to remove it off their site.
|
{
"evidence": [
"Lexulous_sentence_0",
"Lexulous_sentence_1",
"Lexulous_sentence_3",
"Lexulous_sentence_16",
"Lexulous_sentence_17",
"Lexulous_sentence_18"
],
"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:\nAn Indian company running Lexulous, formerly known as Scrabulous was threatened to be shut down before Facebook was asked to remove it off their site.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Lexulous_sentence_0",
"Lexulous_sentence_1",
"Lexulous_sentence_3",
"Lexulous_sentence_16",
"Lexulous_sentence_17",
"Lexulous_sentence_18"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5805,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.594",
"value": "?search="
},
{
"operation": "search",
"time": "14.927",
"value": "Lexulous"
},
{
"operation": "Now on",
"time": "15.781",
"value": "Lexulous"
},
{
"operation": "Highlighting",
"time": "21.38",
"value": "Lexulous_sentence_0"
},
{
"operation": "Highlighting",
"time": "25.548",
"value": "Lexulous_sentence_1"
},
{
"operation": "Highlighting",
"time": "69.363",
"value": "Lexulous_sentence_3"
},
{
"operation": "Highlighting",
"time": "88.437",
"value": "Lexulous_sentence_16"
},
{
"operation": "Highlighting",
"time": "89.198",
"value": "Lexulous_sentence_17"
},
{
"operation": "Highlighting",
"time": "90.22",
"value": "Lexulous_sentence_18"
},
{
"operation": "finish",
"time": "130.152",
"value": "finish"
}
]
}
|
exfever_6038
|
The Midwestern United States is the most populous among the four census regions of the United States.
|
{
"evidence": [
"Midwestern United States",
"United States Census Bureau"
],
"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 Midwestern United States is the most populous among the four census regions of the United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Midwestern United States",
"United States Census Bureau"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5806,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_517
|
The novel by Harper Lee, To Kill a Mockingbird, has become a classic of modern American literature, winning an award for achievements in newspaper, magazine, online journalism, literature, and musical composition within the United Kingdom.
|
{
"evidence": [
"To Kill a Mockingbird",
"Pulitzer Prize"
],
"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 novel by Harper Lee, To Kill a Mockingbird, has become a classic of modern American literature, winning an award for achievements in newspaper, magazine, online journalism, literature, and musical composition within the United Kingdom.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"To Kill a Mockingbird",
"Pulitzer Prize"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5807,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
44311
|
Gaurav Wadhwa is an Indian television actor and younger brother of actor Nirbhay Wadhwa who runs his own YouTube channel, RiMoRav Vlogs.
|
{
"evidence": [
"Gaurav Wadhwa_sentence_3",
"Gaurav Wadhwa_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:\nGaurav Wadhwa is an Indian television actor and younger brother of actor Nirbhay Wadhwa who runs his own YouTube channel, RiMoRav Vlogs.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Gaurav Wadhwa_sentence_3",
"Gaurav Wadhwa_sentence_4"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5808,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.601",
"value": "?search="
},
{
"operation": "search",
"time": "19.243",
"value": "Gaurav Wadhwa"
},
{
"operation": "Now on",
"time": "19.961",
"value": "Gaurav Wadhwa"
},
{
"operation": "Highlighting",
"time": "23.936",
"value": "Gaurav Wadhwa_sentence_0"
},
{
"operation": "Highlighting",
"time": "31.135",
"value": "Gaurav Wadhwa_sentence_6"
},
{
"operation": "Highlighting",
"time": "49.039",
"value": "Gaurav Wadhwa_sentence_3"
},
{
"operation": "Highlighting",
"time": "87.758",
"value": "Gaurav Wadhwa_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "95.366",
"value": "Gaurav Wadhwa_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "104.702",
"value": "Gaurav Wadhwa_sentence_6"
},
{
"operation": "Highlighting deleted",
"time": "107.19",
"value": "Gaurav Wadhwa_sentence_0"
},
{
"operation": "Highlighting",
"time": "119.102",
"value": "Gaurav Wadhwa_sentence_4"
},
{
"operation": "finish",
"time": "129.022",
"value": "finish"
}
]
}
|
6300
|
Knowltonia vesicatoria (of the Ranunculaceae a.k.a. buttercup or crowfoot family) is one of at least ten species of Knowltonia that are native to South America, which can be differentiated from the other species by looking at its leaves and stems being hairless and slightly jagged, toothed margins of the leaves.
|
{
"evidence": [
"Knowltonia vesicatoria_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:\nKnowltonia vesicatoria (of the Ranunculaceae a.k.a. buttercup or crowfoot family) is one of at least ten species of Knowltonia that are native to South America, which can be differentiated from the other species by looking at its leaves and stems being hairless and slightly jagged, toothed margins of the leaves.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Knowltonia vesicatoria_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5809,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.747",
"value": "?search="
},
{
"operation": "search",
"time": "23.975",
"value": "Knowltonia vesicatoria"
},
{
"operation": "Now on",
"time": "24.653",
"value": "Knowltonia vesicatoria"
},
{
"operation": "Highlighting",
"time": "34.687",
"value": "Knowltonia vesicatoria_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "35.629",
"value": "Knowltonia vesicatoria_cell_0_6_0"
},
{
"operation": "Highlighting",
"time": "35.633",
"value": "Knowltonia vesicatoria_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "35.635",
"value": "Knowltonia vesicatoria_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "36.261",
"value": "Knowltonia vesicatoria_cell_0_6_1"
},
{
"operation": "hyperlink",
"time": "37.069",
"value": "Ranunculaceae"
},
{
"operation": "Now on",
"time": "37.641",
"value": "Ranunculaceae"
},
{
"operation": "Highlighting",
"time": "42.989",
"value": "Ranunculaceae_sentence_0"
},
{
"operation": "Highlighting",
"time": "42.993",
"value": "Ranunculaceae_title"
},
{
"operation": "back-button-clicked",
"time": "45.134",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "45.188",
"value": "Knowltonia vesicatoria"
},
{
"operation": "Highlighting",
"time": "54.605",
"value": "Knowltonia vesicatoria_sentence_1"
},
{
"operation": "finish",
"time": "99.756",
"value": "finish"
}
]
}
|
39213
|
The New Village Press, a non-profit book publisher, publishes books addressing topics such as social justice and urban ecology.
|
{
"evidence": [
"New Village Press_sentence_0",
"New Village Press_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:\nThe New Village Press, a non-profit book publisher, publishes books addressing topics such as social justice and urban ecology.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"New Village Press_sentence_0",
"New Village Press_sentence_2"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5810,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.901",
"value": "?search="
},
{
"operation": "search",
"time": "15.269",
"value": "New Village Press"
},
{
"operation": "Now on",
"time": "16.033",
"value": "New Village Press"
},
{
"operation": "Highlighting",
"time": "22.931",
"value": "New Village Press_sentence_0"
},
{
"operation": "Highlighting",
"time": "48.623",
"value": "New Village Press_sentence_2"
},
{
"operation": "finish",
"time": "56.394",
"value": "finish"
}
]
}
|
21060
|
Glen Wallace, born on May 5, 1976 in Ballymena, County Antrim, Southern Ireland, is an actor known for Hollyoaks.
|
{
"evidence": [
"Glen Wallace_sentence_0",
"Glen Wallace_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:\nGlen Wallace, born on May 5, 1976 in Ballymena, County Antrim, Southern Ireland, is an actor known for Hollyoaks.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Glen Wallace_sentence_0",
"Glen Wallace_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5811,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.723",
"value": "?search="
},
{
"operation": "search",
"time": "15.403",
"value": "Glen Wallace"
},
{
"operation": "Now on",
"time": "16.249",
"value": "Glen Wallace"
},
{
"operation": "Highlighting",
"time": "23.921",
"value": "Glen Wallace_sentence_0"
},
{
"operation": "Highlighting",
"time": "24.823",
"value": "Glen Wallace_sentence_1"
},
{
"operation": "finish",
"time": "49.029",
"value": "finish"
}
]
}
|
35001
|
Caguas (a historic place in central Puerto Rico) had seven listings on the registry of Historic Places, and five of these listings are found in a barrio called Caguas Pueblo.
|
{
"evidence": [
"Caguas, Puerto Rico_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:\nCaguas (a historic place in central Puerto Rico) had seven listings on the registry of Historic Places, and five of these listings are found in a barrio called Caguas Pueblo.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Caguas, Puerto Rico_sentence_1"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5812,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.308",
"value": "?search="
},
{
"operation": "search",
"time": "9.507",
"value": "Caguas"
},
{
"operation": "Now on",
"time": "11.545",
"value": "Caguas, Puerto Rico"
},
{
"operation": "Highlighting",
"time": "34.419",
"value": "Caguas, Puerto Rico_sentence_1"
},
{
"operation": "search",
"time": "57.563",
"value": "Caguas, Puerto Rico"
},
{
"operation": "Now on",
"time": "58.928",
"value": "Caguas, Puerto Rico"
},
{
"operation": "hyperlink",
"time": "93.124",
"value": "1.1 Hurricane Maria Caguas, Puerto Rico_section_1"
},
{
"operation": "finish",
"time": "117.344",
"value": "finish"
}
]
}
|
13335
|
Pola Kinski's half siblings Nastassja Kinski and Nikolai Kinski also acted although they barely saw each other while growing up.
|
{
"evidence": [
"Pola Kinski_sentence_7",
"Pola Kinski_sentence_8"
],
"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:\nPola Kinski's half siblings Nastassja Kinski and Nikolai Kinski also acted although they barely saw each other while growing up.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Pola Kinski_sentence_7",
"Pola Kinski_sentence_8"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5813,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.692",
"value": "?search="
},
{
"operation": "search",
"time": "11.568",
"value": "Pola Kinski"
},
{
"operation": "Now on",
"time": "12.621",
"value": "Pola Kinski"
},
{
"operation": "Highlighting",
"time": "33.609",
"value": "Pola Kinski_sentence_7"
},
{
"operation": "Highlighting",
"time": "35.553",
"value": "Pola Kinski_sentence_8"
},
{
"operation": "finish",
"time": "63.859",
"value": "finish"
}
]
}
|
59783
|
Genia Averbuch designed apartment buildings and urban villas in Tel Aviv in 1924, a city that was undergoing a building spurt due to an immigration wave.
|
{
"evidence": [
"Genia Averbuch_sentence_6"
],
"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:\nGenia Averbuch designed apartment buildings and urban villas in Tel Aviv in 1924, a city that was undergoing a building spurt due to an immigration wave.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Genia Averbuch_sentence_6"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5814,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.281",
"value": "?search="
},
{
"operation": "search",
"time": "11.819",
"value": "Genia Averbuch"
},
{
"operation": "Now on",
"time": "12.738",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "14.676",
"value": "Genia Averbuch"
},
{
"operation": "Now on",
"time": "15.552",
"value": "Genia Averbuch"
},
{
"operation": "Page search",
"time": "35.138",
"value": "build"
},
{
"operation": "Page search",
"time": "35.306",
"value": "build"
},
{
"operation": "Highlighting",
"time": "42.677",
"value": "Genia Averbuch_sentence_12"
},
{
"operation": "Highlighting",
"time": "120.935",
"value": "Genia Averbuch_sentence_6"
},
{
"operation": "Highlighting deleted",
"time": "128.597",
"value": "Genia Averbuch_sentence_12"
},
{
"operation": "finish",
"time": "147.235",
"value": "finish"
}
]
}
|
11249
|
Daxiang District, is wrongly believed to cover 214.66 km and is part of the Hunan province in China.
|
{
"evidence": [
"Daxiang District_sentence_0",
"Daxiang District_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:\nDaxiang District, is wrongly believed to cover 214.66 km and is part of the Hunan province in China.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Daxiang District_sentence_0",
"Daxiang District_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5815,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.276",
"value": "?search="
},
{
"operation": "search",
"time": "14.481",
"value": "Daxiang District"
},
{
"operation": "Now on",
"time": "15.807",
"value": "Daxiang District"
},
{
"operation": "Highlighting",
"time": "20.158",
"value": "Daxiang District_sentence_0"
},
{
"operation": "Highlighting",
"time": "34.265",
"value": "Daxiang District_sentence_2"
},
{
"operation": "finish",
"time": "59.635",
"value": "finish"
}
]
}
|
30076
|
Abraham Fischer became the vice-president of the Volksraad in 1893 and headed a joint deputation from Transvaal independent and internationally recognized state in what is now South Africa, from 1852 to 1902.
|
{
"evidence": [
"Abraham Fischer_sentence_7",
"Abraham Fischer_sentence_8",
"South African Republic_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:\nAbraham Fischer became the vice-president of the Volksraad in 1893 and headed a joint deputation from Transvaal independent and internationally recognized state in what is now South Africa, from 1852 to 1902.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Abraham Fischer_sentence_7",
"Abraham Fischer_sentence_8",
"South African Republic_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5816,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.131",
"value": "?search="
},
{
"operation": "search",
"time": "12.611",
"value": "Abraham Fischer"
},
{
"operation": "Now on",
"time": "13.527",
"value": "Abraham Fischer"
},
{
"operation": "Page search",
"time": "20.083",
"value": "1893"
},
{
"operation": "Highlighting",
"time": "23.889",
"value": "Abraham Fischer_sentence_7"
},
{
"operation": "Highlighting",
"time": "29.244",
"value": "Abraham Fischer_sentence_8"
},
{
"operation": "hyperlink",
"time": "72.106",
"value": "Transvaal"
},
{
"operation": "Now on",
"time": "72.792",
"value": "South African Republic"
},
{
"operation": "Highlighting",
"time": "77.314",
"value": "South African Republic_sentence_1"
},
{
"operation": "finish",
"time": "97.114",
"value": "finish"
}
]
}
|
66410
|
While having medium wave transmitters such as 558 kHz (Panjab Radio), and 720 kHz (BBC Radio 4), the Crystal Palace transmitting station was also previously used for FM radio transmission of local radio stations such as Radio X (launched nationally on 21 September 2015 as a rebrand of Xfm), and Capital Xtra, and has been one of five London transmitters for the BBC DAB multiplex since 1995.
|
{
"evidence": [
"Crystal Palace transmitting station_sentence_20",
"Crystal Palace transmitting station_sentence_19",
"Crystal Palace transmitting station_sentence_22",
"Radio X (United Kingdom)_sentence_4"
],
"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:\nWhile having medium wave transmitters such as 558 kHz (Panjab Radio), and 720 kHz (BBC Radio 4), the Crystal Palace transmitting station was also previously used for FM radio transmission of local radio stations such as Radio X (launched nationally on 21 September 2015 as a rebrand of Xfm), and Capital Xtra, and has been one of five London transmitters for the BBC DAB multiplex since 1995.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Crystal Palace transmitting station_sentence_20",
"Crystal Palace transmitting station_sentence_19",
"Crystal Palace transmitting station_sentence_22",
"Radio X (United Kingdom)_sentence_4"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5817,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.132",
"value": "?search="
},
{
"operation": "search",
"time": "10.193",
"value": "While having medium wave transmitters such as 558 kHz (Panjab Radio)"
},
{
"operation": "Now on",
"time": "12.947",
"value": "undefined"
},
{
"operation": "search",
"time": "23.568",
"value": "While having medium wave transmitters such as 558 kHz (Panjab Radio)"
},
{
"operation": "Now on",
"time": "24.618",
"value": "undefined"
},
{
"operation": "search",
"time": "34.215",
"value": "Crystal Palace transmitting station"
},
{
"operation": "Now on",
"time": "35.551",
"value": "Crystal Palace transmitting station"
},
{
"operation": "Page search",
"time": "41.437",
"value": "558"
},
{
"operation": "Highlighting",
"time": "46.493",
"value": "Crystal Palace transmitting station_sentence_20"
},
{
"operation": "Highlighting",
"time": "56.533",
"value": "Crystal Palace transmitting station_sentence_19"
},
{
"operation": "Highlighting",
"time": "75.626",
"value": "Crystal Palace transmitting station_sentence_22"
},
{
"operation": "hyperlink",
"time": "78.158",
"value": "Capital Xtra"
},
{
"operation": "Now on",
"time": "83.398",
"value": "Capital Xtra"
},
{
"operation": "Page search",
"time": "112.174",
"value": "launch"
},
{
"operation": "back-button-clicked",
"time": "164.987",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "165.246",
"value": "Crystal Palace transmitting station"
},
{
"operation": "hyperlink",
"time": "166.682",
"value": "Radio X"
},
{
"operation": "Now on",
"time": "169.904",
"value": "Radio X (United Kingdom)"
},
{
"operation": "Highlighting",
"time": "172.793",
"value": "Radio X (United Kingdom)_sentence_4"
},
{
"operation": "back-button-clicked",
"time": "175.336",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "175.462",
"value": "Crystal Palace transmitting station"
},
{
"operation": "hyperlink",
"time": "179.48",
"value": "Capital Xtra"
},
{
"operation": "Now on",
"time": "179.929",
"value": "Capital Xtra"
},
{
"operation": "finish",
"time": "186.189",
"value": "finish"
}
]
}
|
68401
|
Huvudstabron is a bridge in Sweden that connects the industrial area Ulvsunda and it was built 1964–67.
|
{
"evidence": [
"Huvudstabron_sentence_0",
"Huvudstabron_sentence_1",
"Huvudstabron_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:\nHuvudstabron is a bridge in Sweden that connects the industrial area Ulvsunda and it was built 1964–67.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Huvudstabron_sentence_0",
"Huvudstabron_sentence_1",
"Huvudstabron_sentence_3"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5818,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.766",
"value": "?search="
},
{
"operation": "search",
"time": "13.71",
"value": "Huvudstabron"
},
{
"operation": "Now on",
"time": "14.578",
"value": "Huvudstabron"
},
{
"operation": "Highlighting",
"time": "19.859",
"value": "Huvudstabron_sentence_0"
},
{
"operation": "Highlighting",
"time": "22.18",
"value": "Huvudstabron_sentence_1"
},
{
"operation": "Highlighting",
"time": "26.481",
"value": "Huvudstabron_sentence_3"
},
{
"operation": "finish",
"time": "34.885",
"value": "finish"
}
]
}
|
exfever_2373
|
Coldplay is a British band that draws strongly from a number of other genres, such as the genre that evolved from the 20th-century folk revival.
|
{
"evidence": [
"Coldplay",
"Rock music",
"Folk 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:\nColdplay is a British band that draws strongly from a number of other genres, such as the genre that evolved from the 20th-century folk revival.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Coldplay",
"Rock music",
"Folk music"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5819,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6109
|
Algeria, officially the People's Democratic Republic of Algeria, is a country in the Maghreb region of Northern Africa, a region encompassing the northern portion of the American continent.
|
{
"evidence": [
"Algeria",
"North Africa"
],
"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:\nAlgeria, officially the People's Democratic Republic of Algeria, is a country in the Maghreb region of Northern Africa, a region encompassing the northern portion of the American continent.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Algeria",
"North Africa"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5820,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5495
|
Superunknown is the third studio album released on March 8, 1994 by an American rock band formed in a seaport city on the West Coast of the United States.
|
{
"evidence": [
"Superunknown",
"Soundgarden",
"Seattle"
],
"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:\nSuperunknown is the third studio album released on March 8, 1994 by an American rock band formed in a seaport city on the West Coast of the United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Superunknown",
"Soundgarden",
"Seattle"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5821,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
79217
|
Islington-City Centre West is a commercial and residential neighbourhood in Etobicoke, Toronto, Ontario, Canada and it exits in Brown's Line.
|
{
"evidence": [
"Islington-City Centre West_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:\nIslington-City Centre West is a commercial and residential neighbourhood in Etobicoke, Toronto, Ontario, Canada and it exits in Brown's Line.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Islington-City Centre West_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5822,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.795",
"value": "?search="
},
{
"operation": "search",
"time": "12.725",
"value": "Islington-City Centre West"
},
{
"operation": "Now on",
"time": "13.396",
"value": "Islington-City Centre West"
},
{
"operation": "Highlighting",
"time": "16.94",
"value": "Islington-City Centre West_sentence_0"
},
{
"operation": "Page search",
"time": "32.356",
"value": "brown"
},
{
"operation": "Page search",
"time": "40.404",
"value": "line"
},
{
"operation": "Page search",
"time": "74.643",
"value": "exit"
},
{
"operation": "Page search",
"time": "102.236",
"value": "Brown's"
},
{
"operation": "hyperlink",
"time": "198.388",
"value": "Bloor–Danforth line"
},
{
"operation": "Now on",
"time": "199.074",
"value": "Bloor–Danforth line"
},
{
"operation": "Page search",
"time": "207.247",
"value": "brown"
},
{
"operation": "search",
"time": "225.982",
"value": "islington"
},
{
"operation": "Now on",
"time": "226.716",
"value": "Islington"
},
{
"operation": "search",
"time": "234.335",
"value": "Islington-City Centre West"
},
{
"operation": "Now on",
"time": "234.819",
"value": "Islington-City Centre West"
},
{
"operation": "search",
"time": "283.836",
"value": "Brown's Line"
},
{
"operation": "Now on",
"time": "285.034",
"value": "List of north–south roads in Toronto"
},
{
"operation": "Page search",
"time": "297.827",
"value": "islington"
},
{
"operation": "finish",
"time": "358.531",
"value": "finish"
}
]
}
|
exfever_1899
|
30 Seconds to Mars, produced by a Canadian music producer and keyboardist, is the debut studio album by American rock band Thirty Seconds to Mars.
|
{
"evidence": [
"30 Seconds to Mars (album)",
"Bob Ezrin"
],
"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:\n30 Seconds to Mars, produced by a Canadian music producer and keyboardist, is the debut studio album by American rock band Thirty Seconds to Mars.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"30 Seconds to Mars (album)",
"Bob Ezrin"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5823,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
18190
|
The German Nicaraguan School has grades up to senior high school and was founded in 1867.
|
{
"evidence": [
"German Nicaraguan School_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:\nThe German Nicaraguan School has grades up to senior high school and was founded in 1867.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"German Nicaraguan School_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5824,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.542",
"value": "?search="
},
{
"operation": "search",
"time": "11.91",
"value": "German Nicaraguan School"
},
{
"operation": "Now on",
"time": "12.628",
"value": "German Nicaraguan School"
},
{
"operation": "Highlighting",
"time": "23.482",
"value": "German Nicaraguan School_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "27.405",
"value": "German Nicaraguan School_sentence_1"
},
{
"operation": "Highlighting",
"time": "28.143",
"value": "German Nicaraguan School_sentence_2"
},
{
"operation": "finish",
"time": "33.224",
"value": "finish"
}
]
}
|
exfever_8270
|
Eragon is an action-fantasy film written by Peter and directed by an American visual effects supervisor and film director who worked on numerous major feature films, including Lemony Snickets.
|
{
"evidence": [
"Eragon (film)",
"Stefen Fangmeier"
],
"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:\nEragon is an action-fantasy film written by Peter and directed by an American visual effects supervisor and film director who worked on numerous major feature films, including Lemony Snickets.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Eragon (film)",
"Stefen Fangmeier"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5825,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1894
|
Boromir is a fictional character appears in the second volume of J. R. R. Tolkien's high fantasy novel The Queen of the Rings.
|
{
"evidence": [
"Boromir",
"The Two Towers"
],
"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:\nBoromir is a fictional character appears in the second volume of J. R. R. Tolkien's high fantasy novel The Queen of the Rings.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Boromir",
"The Two Towers"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5826,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
57667
|
Zombie Tycoon was developed for the Playstation Portable, a seventh generation console that competed with Nintendo Switch.
|
{
"evidence": [
"PlayStation Portable_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:\nZombie Tycoon was developed for the Playstation Portable, a seventh generation console that competed with Nintendo Switch.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"PlayStation Portable_sentence_4"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5827,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.884",
"value": "?search="
},
{
"operation": "search",
"time": "11.17",
"value": "Zombie Tycoon"
},
{
"operation": "Now on",
"time": "13.604",
"value": "Zombie Tycoon"
},
{
"operation": "Highlighting",
"time": "21.535",
"value": "Zombie Tycoon_sentence_0"
},
{
"operation": "Highlighting",
"time": "21.536",
"value": "Zombie Tycoon_title"
},
{
"operation": "hyperlink",
"time": "41.046",
"value": "PlayStation Portable"
},
{
"operation": "Now on",
"time": "42.628",
"value": "PlayStation Portable"
},
{
"operation": "Highlighting",
"time": "47.319",
"value": "PlayStation Portable_sentence_4"
},
{
"operation": "Highlighting",
"time": "65.287",
"value": "PlayStation Portable_cell_0_5_1"
},
{
"operation": "Highlighting",
"time": "65.303",
"value": "PlayStation Portable_header_cell_0_5_0"
},
{
"operation": "Highlighting",
"time": "71.974",
"value": "PlayStation Portable_cell_0_4_1"
},
{
"operation": "Highlighting",
"time": "71.982",
"value": "PlayStation Portable_header_cell_0_4_0"
},
{
"operation": "Highlighting",
"time": "95.71",
"value": "PlayStation Portable_sentence_4"
},
{
"operation": "Highlighting",
"time": "118.247",
"value": "PlayStation Portable_sentence_4"
},
{
"operation": "finish",
"time": "124.446",
"value": "finish"
}
]
}
|
exfever_5353
|
Grunge is an alternative rock genre that emerged during the mid-1980s in the American Pacific Northwest state of a state in Pacific Northwest region of the region comprising the westernmost states of the United States.
|
{
"evidence": [
"Grunge",
"Washington (state)",
"Western United States"
],
"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:\nGrunge is an alternative rock genre that emerged during the mid-1980s in the American Pacific Northwest state of a state in Pacific Northwest region of the region comprising the westernmost states of the United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Grunge",
"Washington (state)",
"Western United States"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5828,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
4120
|
Saint Roch was said to be a member of the house of la Croix de Castries, which was ennobled at the end of the 15th century.
|
{
"evidence": [
"House of Castries_sentence_3",
"House of Castries_sentence_4"
],
"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:\nSaint Roch was said to be a member of the house of la Croix de Castries, which was ennobled at the end of the 15th century.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"House of Castries_sentence_3",
"House of Castries_sentence_4"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5829,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "76.75",
"value": "Contains...la Croix de Castries"
},
{
"operation": "search",
"time": "91.423",
"value": "Contains...house of la Croix de Castries"
},
{
"operation": "hyperlink",
"time": "94.269",
"value": "House of Castries"
},
{
"operation": "Page search",
"time": "103.358",
"value": "roch"
},
{
"operation": "Highlighting",
"time": "124.013",
"value": "House of Castries_sentence_3"
},
{
"operation": "Highlighting",
"time": "126.73",
"value": "House of Castries_sentence_4"
},
{
"operation": "finish",
"time": "169.304",
"value": "finish"
}
]
}
|
exfever_5815
|
The Prince of Egypt is a 1998 American animated musical drama film produced by DreamWorks Animation adapted from the second book of the Bible.
|
{
"evidence": [
"The Prince of Egypt",
"Book of Exodus"
],
"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 Prince of Egypt is a 1998 American animated musical drama film produced by DreamWorks Animation adapted from the second book of the Bible.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Prince of Egypt",
"Book of Exodus"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5830,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6438
|
Maggie Smith was the most successful English actress.
|
{
"evidence": [
"Maggie Smith",
"Triple Crown of Acting"
],
"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:\nMaggie Smith was the most successful English actress.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Maggie Smith",
"Triple Crown of Acting"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5831,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
49777
|
Landric is a supporter of Brunhilda, (the queen consort of Austrasia, part of Francia, by marriage to the Merovingian king Sigebert I of Austrasia) and in 604, he is recorded as the tenth Roman Emperor in Frank Dynasty.
|
{
"evidence": [
"Landric_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:\nLandric is a supporter of Brunhilda, (the queen consort of Austrasia, part of Francia, by marriage to the Merovingian king Sigebert I of Austrasia) and in 604, he is recorded as the tenth Roman Emperor in Frank Dynasty.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Landric_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5832,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.854",
"value": "?search="
},
{
"operation": "search",
"time": "20.248",
"value": "Brunhilda of Austrasia"
},
{
"operation": "Now on",
"time": "21.023",
"value": "Brunhilda of Austrasia"
},
{
"operation": "Highlighting",
"time": "27.968",
"value": "Brunhilda of Austrasia_sentence_2"
},
{
"operation": "Highlighting",
"time": "27.968",
"value": "Brunhilda of Austrasia_title"
},
{
"operation": "Page search",
"time": "53.272",
"value": "landric"
},
{
"operation": "hyperlink",
"time": "64.575",
"value": "Landric"
},
{
"operation": "Now on",
"time": "65.097",
"value": "Landric"
},
{
"operation": "Highlighting",
"time": "75.534",
"value": "Landric_sentence_3"
},
{
"operation": "hyperlink",
"time": "91.95",
"value": "Fredegund"
},
{
"operation": "Now on",
"time": "92.344",
"value": "Fredegund"
},
{
"operation": "Page search",
"time": "108.903",
"value": "landric"
},
{
"operation": "finish",
"time": "135.47",
"value": "finish"
}
]
}
|
30763
|
Khursheed Kamal Aziz is an Indian Agriculturist and a coach born in 1927.
|
{
"evidence": [
"Khursheed Kamal Aziz_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:\nKhursheed Kamal Aziz is an Indian Agriculturist and a coach born in 1927.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Khursheed Kamal Aziz_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5833,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.554",
"value": "?search="
},
{
"operation": "search",
"time": "11.399",
"value": "Khursheed Kamal Aziz"
},
{
"operation": "Now on",
"time": "12.288",
"value": "Khursheed Kamal Aziz"
},
{
"operation": "Highlighting",
"time": "23.616",
"value": "Khursheed Kamal Aziz_sentence_0"
},
{
"operation": "Now on",
"time": "43.907",
"value": "Khursheed Kamal Aziz"
},
{
"operation": "finish",
"time": "53.712",
"value": "finish"
}
]
}
|
exfever_5927
|
Kaminey is a 2009 Indian Hindi-language action film featuring an Indian actress, model, and singer.
|
{
"evidence": [
"Kaminey",
"Priyanka Chopra"
],
"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:\nKaminey is a 2009 Indian Hindi-language action film featuring an Indian actress, model, and singer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Kaminey",
"Priyanka Chopra"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5834,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
26397
|
A Distributed Bragg Reflector laser has diffraction grating (which splits and diffracts light into several beams travelling in different directions) on one or both end mirrors, which reflect a narrow band of light back in the cavity which help them be more spectrally stable.
|
{
"evidence": [
"Diffraction grating_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:\nA Distributed Bragg Reflector laser has diffraction grating (which splits and diffracts light into several beams travelling in different directions) on one or both end mirrors, which reflect a narrow band of light back in the cavity which help them be more spectrally stable.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Diffraction grating_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5835,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.602",
"value": "?search="
},
{
"operation": "search",
"time": "25.256",
"value": "Distributed Bragg reflector laser"
},
{
"operation": "Now on",
"time": "26.253",
"value": "Distributed Bragg reflector laser"
},
{
"operation": "search",
"time": "56.206",
"value": "grating"
},
{
"operation": "Now on",
"time": "57.574",
"value": "Grating"
},
{
"operation": "back-button-clicked",
"time": "59.459",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "59.497",
"value": "Distributed Bragg reflector laser"
},
{
"operation": "Page search",
"time": "62.924",
"value": "grating"
},
{
"operation": "search",
"time": "116.74",
"value": "Diffraction grating"
},
{
"operation": "Now on",
"time": "118.071",
"value": "Diffraction grating"
},
{
"operation": "Highlighting",
"time": "126.22",
"value": "Diffraction grating_sentence_0"
},
{
"operation": "search",
"time": "188.645",
"value": "Bragg reflector"
},
{
"operation": "Now on",
"time": "189.652",
"value": "Distributed Bragg reflector"
},
{
"operation": "hyperlink",
"time": "267.645",
"value": "distributed bragg reflector (DBR) lasers"
},
{
"operation": "Now on",
"time": "268.608",
"value": "Distributed Bragg reflector laser"
},
{
"operation": "Page search",
"time": "355.588",
"value": "end mirrors"
},
{
"operation": "hyperlink",
"time": "378.131",
"value": "distributed Bragg reflector"
},
{
"operation": "Now on",
"time": "379.472",
"value": "Distributed Bragg reflector"
},
{
"operation": "finish",
"time": "463.558",
"value": "finish"
}
]
}
|
exfever_6854
|
Katie Price was the first media personality to build a modern castle in Dorset, England.
|
{
"evidence": [
"Katie Price",
"The Sun (United Kingdom)",
"Broadsheet"
],
"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:\nKatie Price was the first media personality to build a modern castle in Dorset, England.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Katie Price",
"The Sun (United Kingdom)",
"Broadsheet"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5836,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8600
|
One World Trade Center is the main building of the rebuilt World Trade Center complex in a state located in the most populous city in the United Kingdom.
|
{
"evidence": [
"One World Trade Center",
"New York City"
],
"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 World Trade Center is the main building of the rebuilt World Trade Center complex in a state located in the most populous city in the United Kingdom.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"One World Trade Center",
"New York City"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5837,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5067
|
The Invention of Lying is a 2009 American romantic comedy film whose cast includes an American actor who appeared in a 1983 American coming-of-age drama film directed by Francis Ford Coppola.
|
{
"evidence": [
"The Invention of Lying",
"Rob Lowe",
"The Outsiders (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:\nThe Invention of Lying is a 2009 American romantic comedy film whose cast includes an American actor who appeared in a 1983 American coming-of-age drama film directed by Francis Ford Coppola.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Invention of Lying",
"Rob Lowe",
"The Outsiders (film)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5838,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
10718
|
European financial institutions could not easily lend to the workers, rendering labor immobile because of capital market imperfections; so to address this imperfection, the Virginia Company would allow laborers to borrow against their future earnings for a fixed number of years.
|
{
"evidence": [
"Plymouth Company_sentence_4",
"Virginia Company_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:\nEuropean financial institutions could not easily lend to the workers, rendering labor immobile because of capital market imperfections; so to address this imperfection, the Virginia Company would allow laborers to borrow against their future earnings for a fixed number of years.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Plymouth Company_sentence_4",
"Virginia Company_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5839,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.244",
"value": "?search="
},
{
"operation": "search",
"time": "25.826",
"value": "Virginia Company"
},
{
"operation": "Now on",
"time": "27.194",
"value": "Virginia Company"
},
{
"operation": "Highlighting",
"time": "44.901",
"value": "Virginia Company_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "55.288",
"value": "Virginia Company_sentence_1"
},
{
"operation": "hyperlink",
"time": "121.446",
"value": "Plymouth Company"
},
{
"operation": "Now on",
"time": "122.629",
"value": "Plymouth Company"
},
{
"operation": "Highlighting",
"time": "130.801",
"value": "Plymouth Company_sentence_4"
},
{
"operation": "hyperlink",
"time": "149.968",
"value": "Virginia Company#Plymouth Company"
},
{
"operation": "Now on",
"time": "150.483",
"value": "Virginia Company#Plymouth Company"
},
{
"operation": "hyperlink",
"time": "155.82",
"value": "London Company"
},
{
"operation": "Now on",
"time": "156.666",
"value": "London Company"
},
{
"operation": "hyperlink",
"time": "257.138",
"value": "Richard Hakluyt"
},
{
"operation": "Now on",
"time": "258.007",
"value": "Richard Hakluyt"
},
{
"operation": "back-button-clicked",
"time": "267.635",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "267.765",
"value": "London Company"
},
{
"operation": "search",
"time": "314.312",
"value": "Virginia Company"
},
{
"operation": "Now on",
"time": "315.216",
"value": "Virginia Company"
},
{
"operation": "Highlighting",
"time": "316.935",
"value": "Virginia Company_sentence_0"
},
{
"operation": "Highlighting",
"time": "324.66",
"value": "Virginia Company_cell_0_0_1"
},
{
"operation": "Highlighting",
"time": "324.663",
"value": "Virginia Company_header_cell_0_0_0"
},
{
"operation": "Highlighting deleted",
"time": "381.221",
"value": "Virginia Company_cell_0_0_1"
},
{
"operation": "Highlighting deleted",
"time": "381.223",
"value": "Virginia Company_header_cell_0_0_0"
},
{
"operation": "hyperlink",
"time": "382.837",
"value": "London Company"
},
{
"operation": "Now on",
"time": "383.775",
"value": "London Company"
},
{
"operation": "finish",
"time": "433.925",
"value": "finish"
}
]
}
|
54493
|
The expression "boots on the ground" has an extended military-jargon history and certainly dates back at least to British officer Robert Grainger Ker Thompson, strategist of the British counter-insurgency efforts against the Malayan National Liberation Army during the Malayan Emergency, a guerrilla war fought in the Federation of Malaya from 1948 until 1960.
|
{
"evidence": [
"Boots on the Ground_sentence_1",
"Boots on the Ground_sentence_2",
"Malayan Emergency_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:\nThe expression \"boots on the ground\" has an extended military-jargon history and certainly dates back at least to British officer Robert Grainger Ker Thompson, strategist of the British counter-insurgency efforts against the Malayan National Liberation Army during the Malayan Emergency, a guerrilla war fought in the Federation of Malaya from 1948 until 1960.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Boots on the Ground_sentence_1",
"Boots on the Ground_sentence_2",
"Malayan Emergency_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5840,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.964",
"value": "?search="
},
{
"operation": "search",
"time": "25.042",
"value": "Boots on the Ground"
},
{
"operation": "Now on",
"time": "25.809",
"value": "Boots on the Ground"
},
{
"operation": "Highlighting",
"time": "31.753",
"value": "Boots on the Ground_sentence_1"
},
{
"operation": "Highlighting",
"time": "42.737",
"value": "Boots on the Ground_sentence_2"
},
{
"operation": "hyperlink",
"time": "60.169",
"value": "Malayan Emergency"
},
{
"operation": "Now on",
"time": "65.357",
"value": "Malayan Emergency"
},
{
"operation": "Highlighting",
"time": "66.588",
"value": "Malayan Emergency_sentence_0"
},
{
"operation": "finish",
"time": "119.368",
"value": "finish"
}
]
}
|
exfever_8171
|
Logan is a 2017 Bristish superhero film starring Hugh Jackman a fictional character appearing in American comic books published by Marvel Comics.
|
{
"evidence": [
"Logan (film)",
"Wolverine (character)"
],
"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:\nLogan is a 2017 Bristish superhero film starring Hugh Jackman a fictional character appearing in American comic books published by Marvel Comics.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Logan (film)",
"Wolverine (character)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5841,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6784
|
NBC was founded in 1926 by the Radio Corporation of America.
|
{
"evidence": [
"JAG (TV series)",
"NBC",
"Comcast"
],
"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:\nNBC was founded in 1926 by the Radio Corporation of America.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"JAG (TV series)",
"NBC",
"Comcast"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5842,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
25541
|
Should the World Fail to Fall Apart is an album by Peter Murphy which will be released on 2022.
|
{
"evidence": [
"Should the World Fail to Fall Apart_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:\nShould the World Fail to Fall Apart is an album by Peter Murphy which will be released on 2022.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Should the World Fail to Fall Apart_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5843,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.556",
"value": "?search="
},
{
"operation": "search",
"time": "18.794",
"value": "Contains...Should the World Fail to Fal"
},
{
"operation": "Now on",
"time": "20.511",
"value": "undefined"
},
{
"operation": "search",
"time": "32.491",
"value": "Peter Murphy"
},
{
"operation": "Now on",
"time": "33.165",
"value": "Peter Murphy"
},
{
"operation": "Highlighting",
"time": "52.261",
"value": "Peter Murphy_item_0_1"
},
{
"operation": "Highlighting",
"time": "52.261",
"value": "Peter Murphy_title"
},
{
"operation": "Highlighting deleted",
"time": "55.591",
"value": "Peter Murphy_item_0_1"
},
{
"operation": "Highlighting deleted",
"time": "55.592",
"value": "Peter Murphy_title"
},
{
"operation": "hyperlink",
"time": "56.896",
"value": "Peter Murphy (artist)"
},
{
"operation": "Now on",
"time": "57.539",
"value": "Peter Murphy (artist)"
},
{
"operation": "search",
"time": "78.988",
"value": "Should the World Fail to Fall Apart"
},
{
"operation": "Now on",
"time": "79.846",
"value": "Should the World Fail to Fall Apart"
},
{
"operation": "Highlighting",
"time": "93.121",
"value": "Should the World Fail to Fall Apart_sentence_2"
},
{
"operation": "finish",
"time": "120.177",
"value": "finish"
}
]
}
|
exfever_2146
|
Splash is an American fantasy romantic comedy film starring an American actress who also starred as Pris Stratton in Blade Runner by an Italian film director and producer.
|
{
"evidence": [
"Splash (film)",
"Daryl Hannah",
"Ridley 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:\nSplash is an American fantasy romantic comedy film starring an American actress who also starred as Pris Stratton in Blade Runner by an Italian film director and producer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Splash (film)",
"Daryl Hannah",
"Ridley Scott"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5844,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
35623
|
Iván Villazón teamed up with flute player Cocha molina who was also a Vallenato Legend Festival winner, one of the most important musical festivals in Colombia, and recorded an album with three major hits.
|
{
"evidence": [
"Iván Villazón_sentence_14"
],
"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:\nIván Villazón teamed up with flute player Cocha molina who was also a Vallenato Legend Festival winner, one of the most important musical festivals in Colombia, and recorded an album with three major hits.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Iván Villazón_sentence_14"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5845,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.716",
"value": "?search="
},
{
"operation": "search",
"time": "14.987",
"value": "Iván Villazón"
},
{
"operation": "Now on",
"time": "15.801",
"value": "Iván Villazón"
},
{
"operation": "Highlighting",
"time": "61.527",
"value": "Iván Villazón_sentence_14"
},
{
"operation": "finish",
"time": "71.677",
"value": "finish"
}
]
}
|
exfever_4284
|
Debie Reynolds got her Golden Globe nomination in the 1956 Technicolor musical remake of an American romantic comedy film directed by Garson Kanin.
|
{
"evidence": [
"Debbie Reynolds",
"Bundle of Joy",
"Bachelor Mother"
],
"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:\nDebie Reynolds got her Golden Globe nomination in the 1956 Technicolor musical remake of an American romantic comedy film directed by Garson Kanin.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Debbie Reynolds",
"Bundle of Joy",
"Bachelor Mother"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5846,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
35233
|
Singapore had the most number of athletes participating in cue sports at the Asian Games in 2010 (held in Guangzhou, China).
|
{
"evidence": [
"Cue sports at the 2010 Asian Games_sentence_1",
"Cue sports at the 2010 Asian Games_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:\nSingapore had the most number of athletes participating in cue sports at the Asian Games in 2010 (held in Guangzhou, China).\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Cue sports at the 2010 Asian Games_sentence_1",
"Cue sports at the 2010 Asian Games_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5847,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.654",
"value": "?search="
},
{
"operation": "search",
"time": "58.133",
"value": "Cue sports at the 2010 Asian Games"
},
{
"operation": "Now on",
"time": "59.145",
"value": "Cue sports at the 2010 Asian Games"
},
{
"operation": "Page search",
"time": "61.328",
"value": "Singapore"
},
{
"operation": "Highlighting",
"time": "84.729",
"value": "Cue sports at the 2010 Asian Games_sentence_1"
},
{
"operation": "Highlighting",
"time": "210.28",
"value": "Cue sports at the 2010 Asian Games_cell_3_4_0"
},
{
"operation": "Highlighting",
"time": "210.281",
"value": "Cue sports at the 2010 Asian Games_section_1"
},
{
"operation": "Highlighting",
"time": "210.281",
"value": "Cue sports at the 2010 Asian Games_section_2"
},
{
"operation": "Highlighting",
"time": "210.282",
"value": "Cue sports at the 2010 Asian Games_header_cell_3_0_3"
},
{
"operation": "Highlighting deleted",
"time": "213.961",
"value": "Cue sports at the 2010 Asian Games_cell_3_4_0"
},
{
"operation": "Highlighting deleted",
"time": "213.962",
"value": "Cue sports at the 2010 Asian Games_section_1"
},
{
"operation": "Highlighting deleted",
"time": "213.963",
"value": "Cue sports at the 2010 Asian Games_section_2"
},
{
"operation": "Highlighting deleted",
"time": "213.964",
"value": "Cue sports at the 2010 Asian Games_header_cell_3_0_3"
},
{
"operation": "Highlighting",
"time": "226.561",
"value": "Cue sports at the 2010 Asian Games_sentence_0"
},
{
"operation": "finish",
"time": "230.547",
"value": "finish"
}
]
}
|
62512
|
The Iris latistyla has two bract (protective layer) and the flowers are 5 cm in diameter with two pairs of petals.
|
{
"evidence": [
"Iris latistyla_sentence_13",
"Iris latistyla_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:\nThe Iris latistyla has two bract (protective layer) and the flowers are 5 cm in diameter with two pairs of petals.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Iris latistyla_sentence_13",
"Iris latistyla_sentence_14"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5848,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.468",
"value": "?search="
},
{
"operation": "search",
"time": "45.24",
"value": "Iris latistyla"
},
{
"operation": "Now on",
"time": "48.569",
"value": "Iris latistyla"
},
{
"operation": "Highlighting",
"time": "154.028",
"value": "Iris latistyla_sentence_13"
},
{
"operation": "Highlighting",
"time": "154.86",
"value": "Iris latistyla_sentence_14"
},
{
"operation": "finish",
"time": "215.869",
"value": "finish"
}
]
}
|
22092
|
Baseball was introduced in the Philippines by the Irish, and it was first played in the Philippines in 1898; the Phillipines National baseball team participated in Far Eastern games every two years from 1913 through 1934.
|
{
"evidence": [
"Baseball in the Philippines_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:\nBaseball was introduced in the Philippines by the Irish, and it was first played in the Philippines in 1898; the Phillipines National baseball team participated in Far Eastern games every two years from 1913 through 1934.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Baseball in the Philippines_sentence_4"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5849,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.124",
"value": "?search="
},
{
"operation": "search",
"time": "14.358",
"value": "Contains...Baseball was introduced in the Philippines by the Irish"
},
{
"operation": "Now on",
"time": "16.611",
"value": "undefined"
},
{
"operation": "search",
"time": "39.773",
"value": "Baseball in the Philippines"
},
{
"operation": "Now on",
"time": "40.339",
"value": "Baseball in the Philippines"
},
{
"operation": "Page search",
"time": "58.163",
"value": "Irish"
},
{
"operation": "Highlighting",
"time": "82.187",
"value": "Baseball in the Philippines_sentence_4"
},
{
"operation": "finish",
"time": "115.994",
"value": "finish"
}
]
}
|
exfever_7570
|
The United States Naval Academy is a federal service academy in Annapolis that educates midshipmen for service in the officer corps of the maritime service branch of the United States Armed Forces.
|
{
"evidence": [
"United States Naval Academy",
"United States Navy"
],
"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 United States Naval Academy is a federal service academy in Annapolis that educates midshipmen for service in the officer corps of the maritime service branch of the United States Armed Forces.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"United States Naval Academy",
"United States Navy"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5850,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
11158
|
Christianity, in its early development, accommodated many traditional practices which at different times, accepted omens and practiced forms of ritual sacrifice; and concepts of "Divine Grace" or "Blessing", as they are described by believers, closely resemble what is referred to as "luck" (the phenomenon and belief that defines the experience of notably positive, negative, or improbable events) by others.
|
{
"evidence": [
"Luck_sentence_2",
"Divine grace_sentence_18"
],
"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:\nChristianity, in its early development, accommodated many traditional practices which at different times, accepted omens and practiced forms of ritual sacrifice; and concepts of \"Divine Grace\" or \"Blessing\", as they are described by believers, closely resemble what is referred to as \"luck\" (the phenomenon and belief that defines the experience of notably positive, negative, or improbable events) by others.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Luck_sentence_2",
"Divine grace_sentence_18"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5851,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.063",
"value": "undefined"
},
{
"operation": "search",
"time": "112.979",
"value": "Contains...\"Divine Grace\""
},
{
"operation": "Now on",
"time": "115.285",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "128.873",
"value": "Merit (Christianity)"
},
{
"operation": "Now on",
"time": "130.556",
"value": "Merit (Christianity)"
},
{
"operation": "search",
"time": "259.772",
"value": "Luck"
},
{
"operation": "Now on",
"time": "261.839",
"value": "Luck"
},
{
"operation": "Highlighting",
"time": "268.02",
"value": "Luck_sentence_2"
},
{
"operation": "hyperlink",
"time": "293.979",
"value": "4 In religion and mythology Luck_section_12"
},
{
"operation": "search",
"time": "340.808",
"value": "sacrifice"
},
{
"operation": "Now on",
"time": "343.369",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "361.885",
"value": "Salvation in Christianity"
},
{
"operation": "Now on",
"time": "363.091",
"value": "Salvation in Christianity"
},
{
"operation": "back-button-clicked",
"time": "415.849",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "416.72",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "422.749",
"value": "Isaac"
},
{
"operation": "Now on",
"time": "423.702",
"value": "Isaac"
},
{
"operation": "hyperlink",
"time": "457.557",
"value": "mercy of God"
},
{
"operation": "Now on",
"time": "458.667",
"value": "Atonement in Judaism"
},
{
"operation": "Highlighting",
"time": "466.215",
"value": "Atonement in Judaism_sentence_3"
},
{
"operation": "Highlighting",
"time": "466.216",
"value": "Atonement in Judaism_title"
},
{
"operation": "back-button-clicked",
"time": "488.751",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "488.835",
"value": "Isaac"
},
{
"operation": "hyperlink",
"time": "514.414",
"value": "Jewish"
},
{
"operation": "Now on",
"time": "516.64",
"value": "Jewish"
},
{
"operation": "hyperlink",
"time": "540.285",
"value": "4.1 Babylon and Rome Jews_section_4"
},
{
"operation": "search",
"time": "581.155",
"value": "divine grace"
},
{
"operation": "Now on",
"time": "582.29",
"value": "Divine grace"
},
{
"operation": "hyperlink",
"time": "585.813",
"value": "2 Christianity Divine grace_section_1"
},
{
"operation": "Highlighting",
"time": "602.352",
"value": "Divine grace_sentence_18"
},
{
"operation": "hyperlink",
"time": "636.716",
"value": "Grace in Christianity"
},
{
"operation": "Now on",
"time": "637.782",
"value": "Grace in Christianity"
},
{
"operation": "back-button-clicked",
"time": "641.596",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "641.692",
"value": "Divine grace#Christianity Divine grace section 1"
},
{
"operation": "search",
"time": "651.304",
"value": "blessing"
},
{
"operation": "Now on",
"time": "652.948",
"value": "Blessing"
},
{
"operation": "search",
"time": "705.955",
"value": "Sacrifice in Christianity"
},
{
"operation": "Now on",
"time": "707.823",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "726.229",
"value": "Early Christianity Christianity_section_24"
},
{
"operation": "Now on",
"time": "727.663",
"value": "Christianity#Early Christianity Christianity section 24"
},
{
"operation": "hyperlink",
"time": "769.16",
"value": "Proto-orthodox Christianity"
},
{
"operation": "Now on",
"time": "770.315",
"value": "Proto-orthodox Christianity"
},
{
"operation": "back-button-clicked",
"time": "790.816",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "791.104",
"value": "Christianity#Early Christianity Christianity section 24"
},
{
"operation": "hyperlink",
"time": "840.408",
"value": "Massacre of Verden"
},
{
"operation": "Now on",
"time": "841.932",
"value": "Massacre of Verden"
},
{
"operation": "back-button-clicked",
"time": "875.523",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "875.711",
"value": "Christianity#Early Christianity Christianity section 24"
},
{
"operation": "finish",
"time": "908.319",
"value": "finish"
}
]
}
|
exfever_300
|
Peter Cetera is an American musician best known for being a lead vocalist and bassist of an American rock band, self-described as a rock and roll band with a family of musical instruments made of a tube.
|
{
"evidence": [
"Peter Cetera",
"Chicago (band)",
"Horn (instrument)"
],
"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:\nPeter Cetera is an American musician best known for being a lead vocalist and bassist of an American rock band, self-described as a rock and roll band with a family of musical instruments made of a tube.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Peter Cetera",
"Chicago (band)",
"Horn (instrument)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5852,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
62045
|
John King of Pollard Ball took accepted the position of chairman of the board at Dennis Specialist Vehicles (an English manufacturer of commercial vehicles based in Guildford), and was joined by Julian Amery, a former Minister of Aviation, as well as a new CEO and a new finance director.
|
{
"evidence": [
"Dennis Specialist Vehicles_sentence_94",
"Dennis Specialist Vehicles_sentence_95",
"Dennis Specialist Vehicles_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:\nJohn King of Pollard Ball took accepted the position of chairman of the board at Dennis Specialist Vehicles (an English manufacturer of commercial vehicles based in Guildford), and was joined by Julian Amery, a former Minister of Aviation, as well as a new CEO and a new finance director.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Dennis Specialist Vehicles_sentence_94",
"Dennis Specialist Vehicles_sentence_95",
"Dennis Specialist Vehicles_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5853,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.712",
"value": "?search="
},
{
"operation": "search",
"time": "13.943",
"value": "Contains...John King of Pollard Ball"
},
{
"operation": "Now on",
"time": "16.442",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "96.964",
"value": "1948 Nationalisation of road haulage Dennis Specialist Vehicles_section_10"
},
{
"operation": "Now on",
"time": "98.293",
"value": "Dennis Specialist Vehicles#1948 Nationalisation of road haulage Dennis Specialist Vehicles section 10"
},
{
"operation": "Highlighting",
"time": "106.893",
"value": "Dennis Specialist Vehicles_sentence_94"
},
{
"operation": "Highlighting",
"time": "114.364",
"value": "Dennis Specialist Vehicles_sentence_95"
},
{
"operation": "Highlighting",
"time": "193.687",
"value": "Dennis Specialist Vehicles_sentence_0"
},
{
"operation": "finish",
"time": "208.43",
"value": "finish"
}
]
}
|
exfever_7205
|
Clark Gable has his own walk of fame star in Hollywood.
|
{
"evidence": [
"Clark Gable",
"It Happened One Night"
],
"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:\nClark Gable has his own walk of fame star in Hollywood.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Clark Gable",
"It Happened One Night"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5854,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
36157
|
Aton Capital took part in the IPO of the Moscow Stock Exchange and was one of the smallest players in the French asset management market.
|
{
"evidence": [
"Aton Capital_sentence_17",
"Aton Capital_sentence_15"
],
"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:\nAton Capital took part in the IPO of the Moscow Stock Exchange and was one of the smallest players in the French asset management market.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Aton Capital_sentence_17",
"Aton Capital_sentence_15"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5855,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.609",
"value": "?search="
},
{
"operation": "Now on",
"time": "26.876",
"value": "?search="
},
{
"operation": "search",
"time": "40.343",
"value": "Aton Capital"
},
{
"operation": "Now on",
"time": "41.083",
"value": "Aton Capital"
},
{
"operation": "Page search",
"time": "79.489",
"value": "french"
},
{
"operation": "Page search",
"time": "85.682",
"value": "player"
},
{
"operation": "Highlighting",
"time": "92.054",
"value": "Aton Capital_sentence_17"
},
{
"operation": "Page search",
"time": "103.578",
"value": "ipo"
},
{
"operation": "Highlighting",
"time": "113.324",
"value": "Aton Capital_sentence_15"
},
{
"operation": "finish",
"time": "131.976",
"value": "finish"
}
]
}
|
48976
|
Galindo, the sighted ski guide for visually impaired skier Jon Santacana, earned a gold medal at the Switzerland hosted 2000 World Championships.
|
{
"evidence": [
"Miguel Galindo Garcés_sentence_5",
"Miguel Galindo Garcés_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:\nGalindo, the sighted ski guide for visually impaired skier Jon Santacana, earned a gold medal at the Switzerland hosted 2000 World Championships.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Miguel Galindo Garcés_sentence_5",
"Miguel Galindo Garcés_sentence_6"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5856,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.758",
"value": "?search="
},
{
"operation": "search",
"time": "7.584",
"value": "Galindo"
},
{
"operation": "Now on",
"time": "8.462",
"value": "Galindo"
},
{
"operation": "search",
"time": "258.718",
"value": "v"
},
{
"operation": "Now on",
"time": "259.725",
"value": "V"
},
{
"operation": "search",
"time": "263.148",
"value": "Jon Santacana"
},
{
"operation": "Now on",
"time": "263.95",
"value": "Jon Santacana Maiztegui"
},
{
"operation": "Highlighting",
"time": "266.966",
"value": "Jon Santacana Maiztegui_sentence_2"
},
{
"operation": "Highlighting",
"time": "266.966",
"value": "Jon Santacana Maiztegui_title"
},
{
"operation": "hyperlink",
"time": "270.744",
"value": "Miguel Galindo Garces"
},
{
"operation": "Now on",
"time": "271.507",
"value": "Miguel Galindo Garcés"
},
{
"operation": "Highlighting",
"time": "286.722",
"value": "Miguel Galindo Garcés_sentence_5"
},
{
"operation": "Highlighting",
"time": "291.233",
"value": "Miguel Galindo Garcés_sentence_6"
},
{
"operation": "finish",
"time": "301.892",
"value": "finish"
}
]
}
|
67286
|
On February 2nd, the FCC (Federal Communications Commission) approves Time Warner Cable's $3 million purchase of Insight Communications, which is a public relations agency providing communications strategy and related services in the Korean market, then on February 3rd, the FCC ruled that television stations have the right to refuse to air ads under "reasonable access" provisions of the law pertaining to political candidates.
|
{
"evidence": [
"Time Warner Cable_sentence_66",
"Time Warner Cable_sentence_67"
],
"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:\nOn February 2nd, the FCC (Federal Communications Commission) approves Time Warner Cable's $3 million purchase of Insight Communications, which is a public relations agency providing communications strategy and related services in the Korean market, then on February 3rd, the FCC ruled that television stations have the right to refuse to air ads under \"reasonable access\" provisions of the law pertaining to political candidates.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Time Warner Cable_sentence_66",
"Time Warner Cable_sentence_67"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5857,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.741",
"value": "?search="
},
{
"operation": "search",
"time": "21.046",
"value": "Contains...approves Time Warner Cable's $3 million purchase of Insight Communications, which is a public relations agency providing communications strategy and related services in the Korean market"
},
{
"operation": "Now on",
"time": "25.283",
"value": "undefined"
},
{
"operation": "search",
"time": "54.961",
"value": "Insight Communications"
},
{
"operation": "Now on",
"time": "56.959",
"value": "Time Warner Cable"
},
{
"operation": "Page search",
"time": "104.941",
"value": "Insight Communications"
},
{
"operation": "Highlighting",
"time": "114.137",
"value": "Time Warner Cable_sentence_66"
},
{
"operation": "Highlighting",
"time": "126.242",
"value": "Time Warner Cable_sentence_67"
},
{
"operation": "search",
"time": "275.374",
"value": "Fcc"
},
{
"operation": "Now on",
"time": "279.292",
"value": "Federal Communications Commission"
},
{
"operation": "Page search",
"time": "284.792",
"value": "Insight Communications"
},
{
"operation": "finish",
"time": "326.534",
"value": "finish"
}
]
}
|
36065
|
The Ammons haunting case occurred in Gary, Indiana where numerous movies have been filmed; and the Department of Child Services was alerted to the Ammon family whose story was sensationalized by news outlets.
|
{
"evidence": [
"Ammons haunting case_sentence_0",
"Ammons haunting case_sentence_41",
"Ammons haunting case_sentence_17",
"Ammons haunting case_sentence_19"
],
"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 Ammons haunting case occurred in Gary, Indiana where numerous movies have been filmed; and the Department of Child Services was alerted to the Ammon family whose story was sensationalized by news outlets.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Ammons haunting case_sentence_0",
"Ammons haunting case_sentence_41",
"Ammons haunting case_sentence_17",
"Ammons haunting case_sentence_19"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5858,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.091",
"value": "?search="
},
{
"operation": "search",
"time": "24.299",
"value": "Ammons haunting case"
},
{
"operation": "Now on",
"time": "25.292",
"value": "Ammons haunting case"
},
{
"operation": "Highlighting",
"time": "31.508",
"value": "Ammons haunting case_sentence_0"
},
{
"operation": "Page search",
"time": "39.32",
"value": "movies"
},
{
"operation": "Page search",
"time": "65.641",
"value": "movie"
},
{
"operation": "Page search",
"time": "74.106",
"value": "films"
},
{
"operation": "Highlighting",
"time": "149.577",
"value": "Ammons haunting case_sentence_41"
},
{
"operation": "Page search",
"time": "171.873",
"value": "child"
},
{
"operation": "Highlighting",
"time": "197.667",
"value": "Ammons haunting case_sentence_17"
},
{
"operation": "Highlighting",
"time": "228.49",
"value": "Ammons haunting case_sentence_19"
},
{
"operation": "finish",
"time": "279.186",
"value": "finish"
}
]
}
|
35944
|
In 1892, Grantown-on-Spey in Strathspeym formed the first shinty club, a club that plays shinty/hurling that appears prominently in the legend of Cúchulainn, the Celtic mythology, and during the late nineteenth century, twelve shinty clubs from across the Badenoch and Strathspey area were in operation.
|
{
"evidence": [
"Strathspey Camanachd_sentence_4",
"Shinty_sentence_41",
"Shinty_sentence_13",
"Grantown-on-Spey_sentence_0",
"Nethy Bridge_sentence_0",
"Boat of Garten_sentence_0",
"Dulnain Bridge_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:\nIn 1892, Grantown-on-Spey in Strathspeym formed the first shinty club, a club that plays shinty/hurling that appears prominently in the legend of Cúchulainn, the Celtic mythology, and during the late nineteenth century, twelve shinty clubs from across the Badenoch and Strathspey area were in operation.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Strathspey Camanachd_sentence_4",
"Shinty_sentence_41",
"Shinty_sentence_13",
"Grantown-on-Spey_sentence_0",
"Nethy Bridge_sentence_0",
"Boat of Garten_sentence_0",
"Dulnain Bridge_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5859,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.621",
"value": "?search="
},
{
"operation": "search",
"time": "31.102",
"value": "Contains...shinty club"
},
{
"operation": "Now on",
"time": "32.069",
"value": "undefined"
},
{
"operation": "search",
"time": "63.717",
"value": "Grantown-on-Spey"
},
{
"operation": "Now on",
"time": "64.431",
"value": "Grantown-on-Spey"
},
{
"operation": "search",
"time": "88.044",
"value": "first shinty club"
},
{
"operation": "Now on",
"time": "89.135",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "105.544",
"value": "History of Shinty in Strathspey Strathspey Camanachd_section_0"
},
{
"operation": "Now on",
"time": "106.689",
"value": "Strathspey Camanachd#History of Shinty in Strathspey Strathspey Camanachd section 0"
},
{
"operation": "hyperlink",
"time": "167.285",
"value": "shinty"
},
{
"operation": "Now on",
"time": "167.926",
"value": "Shinty"
},
{
"operation": "Highlighting",
"time": "205.468",
"value": "Shinty_sentence_41"
},
{
"operation": "search",
"time": "268.468",
"value": "Grantown-on-Spey"
},
{
"operation": "Now on",
"time": "269.189",
"value": "Grantown-on-Spey"
},
{
"operation": "search",
"time": "301.886",
"value": "Contains...Strathspeym"
},
{
"operation": "Now on",
"time": "302.916",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "314.06",
"value": "Strathspey"
},
{
"operation": "Now on",
"time": "314.799",
"value": "Strathspey"
},
{
"operation": "hyperlink",
"time": "328.725",
"value": "Strathspey Camanachd"
},
{
"operation": "Now on",
"time": "329.161",
"value": "Strathspey Camanachd"
},
{
"operation": "Highlighting",
"time": "343.246",
"value": "Strathspey Camanachd_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "429.47",
"value": "Strathspey Camanachd_sentence_1"
},
{
"operation": "Highlighting",
"time": "440.146",
"value": "Strathspey Camanachd_sentence_4"
},
{
"operation": "search",
"time": "498.229",
"value": "Shinty"
},
{
"operation": "Now on",
"time": "499.177",
"value": "Shinty"
},
{
"operation": "Highlighting",
"time": "576.713",
"value": "Shinty_sentence_41"
},
{
"operation": "Highlighting",
"time": "595.93",
"value": "Shinty_sentence_13"
},
{
"operation": "Page search",
"time": "638.936",
"value": "Badenoch"
},
{
"operation": "hyperlink",
"time": "707.545",
"value": "Shinty league system"
},
{
"operation": "Now on",
"time": "708.065",
"value": "Shinty league system"
},
{
"operation": "hyperlink",
"time": "744.565",
"value": "shinty"
},
{
"operation": "Now on",
"time": "745.037",
"value": "Shinty"
},
{
"operation": "search",
"time": "774.973",
"value": "Grantown-on-Spey"
},
{
"operation": "Now on",
"time": "775.696",
"value": "Grantown-on-Spey"
},
{
"operation": "Highlighting",
"time": "800.334",
"value": "Grantown-on-Spey_sentence_0"
},
{
"operation": "search",
"time": "811.054",
"value": "Strathspey"
},
{
"operation": "Now on",
"time": "811.958",
"value": "Strathspey"
},
{
"operation": "search",
"time": "850.949",
"value": "nineteenth century shinty"
},
{
"operation": "Now on",
"time": "852.167",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "863.558",
"value": "History of Shinty in Strathspey Strathspey Camanachd_section_0"
},
{
"operation": "Now on",
"time": "864.304",
"value": "Strathspey Camanachd#History of Shinty in Strathspey Strathspey Camanachd section 0"
},
{
"operation": "hyperlink",
"time": "902.648",
"value": "Nethy Bridge"
},
{
"operation": "Now on",
"time": "903.17",
"value": "Nethy Bridge"
},
{
"operation": "Highlighting",
"time": "908.386",
"value": "Nethy Bridge_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "910.524",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "910.66",
"value": "Strathspey Camanachd#History of Shinty in Strathspey Strathspey Camanachd section 0"
},
{
"operation": "hyperlink",
"time": "916.445",
"value": "Boat of Garten"
},
{
"operation": "Now on",
"time": "916.993",
"value": "Boat of Garten"
},
{
"operation": "Highlighting",
"time": "919.543",
"value": "Boat of Garten_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "921.76",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "921.895",
"value": "Strathspey Camanachd#History of Shinty in Strathspey Strathspey Camanachd section 0"
},
{
"operation": "hyperlink",
"time": "926.151",
"value": "Dulnain Bridge"
},
{
"operation": "Now on",
"time": "926.84",
"value": "Dulnain Bridge"
},
{
"operation": "Highlighting",
"time": "934.175",
"value": "Dulnain Bridge_sentence_0"
},
{
"operation": "finish",
"time": "951.165",
"value": "finish"
}
]
}
|
exfever_2619
|
How to Train Your Dragon is a 2010 American computer-animated action-fantasy film directed by a Canadian film director who is best known for directing the Oscar-nominated animated films Lilo & Stitch for Walt Disney Animation Studios.
|
{
"evidence": [
"How to Train Your Dragon (film)",
"Dean DeBlois"
],
"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:\nHow to Train Your Dragon is a 2010 American computer-animated action-fantasy film directed by a Canadian film director who is best known for directing the Oscar-nominated animated films Lilo & Stitch for Walt Disney Animation Studios.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"How to Train Your Dragon (film)",
"Dean DeBlois"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5860,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6265
|
The Penny Dreadful is a horror drama television series produced by a British film director who directed the play The Ferryman, for which he won the award, which has been given since 1960.
|
{
"evidence": [
"Penny Dreadful (TV series)",
"Sam Mendes",
"Tony Award for Best Direction of a Play"
],
"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 Penny Dreadful is a horror drama television series produced by a British film director who directed the play The Ferryman, for which he won the award, which has been given since 1960.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Penny Dreadful (TV series)",
"Sam Mendes",
"Tony Award for Best Direction of a Play"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5861,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
27949
|
Guru Kelu Nair remained as a faculty member in Santiniketan until the death of Tagore in 1941, and has noteworthy disciples including Mrinalini Sarabhai (the founder and director of the Darpana Academy of Performing Arts), Rukmini Devi Arundale, and Yog Sunder Desai.
|
{
"evidence": [
"Guru Kelu Nair_sentence_6",
"Guru Kelu Nair_sentence_7",
"Mrinalini Sarabhai_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:\nGuru Kelu Nair remained as a faculty member in Santiniketan until the death of Tagore in 1941, and has noteworthy disciples including Mrinalini Sarabhai (the founder and director of the Darpana Academy of Performing Arts), Rukmini Devi Arundale, and Yog Sunder Desai.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Guru Kelu Nair_sentence_6",
"Guru Kelu Nair_sentence_7",
"Mrinalini Sarabhai_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5862,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.041",
"value": "?search="
},
{
"operation": "search",
"time": "7.232",
"value": "Guru Kelu Nair"
},
{
"operation": "Now on",
"time": "8.008",
"value": "Guru Kelu Nair"
},
{
"operation": "Highlighting",
"time": "14.053",
"value": "Guru Kelu Nair_sentence_6"
},
{
"operation": "Highlighting",
"time": "21.511",
"value": "Guru Kelu Nair_sentence_7"
},
{
"operation": "hyperlink",
"time": "31.647",
"value": "Mrinalini Sarabhai"
},
{
"operation": "Now on",
"time": "32.177",
"value": "Mrinalini Sarabhai"
},
{
"operation": "Highlighting",
"time": "38.071",
"value": "Mrinalini Sarabhai_sentence_1"
},
{
"operation": "finish",
"time": "51.953",
"value": "finish"
}
]
}
|
exfever_3368
|
Grunge is an alternative rock genre that fuses elements of punk rock and a genre of rock music characterized by extended melodic passages, instrumental sections, or an entire piece of music that is pre-written to be played on an electric drum.
|
{
"evidence": [
"Grunge",
"Heavy metal music",
"Guitar solo"
],
"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:\nGrunge is an alternative rock genre that fuses elements of punk rock and a genre of rock music characterized by extended melodic passages, instrumental sections, or an entire piece of music that is pre-written to be played on an electric drum.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Grunge",
"Heavy metal music",
"Guitar solo"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5863,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
69474
|
Theseia athletic events were conducted by hoplites in their armor and the first events were only open to Athenian males.
|
{
"evidence": [
"Theseia_sentence_4",
"Theseia_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:\nTheseia athletic events were conducted by hoplites in their armor and the first events were only open to Athenian males.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Theseia_sentence_4",
"Theseia_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5864,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.674",
"value": "undefined"
},
{
"operation": "search",
"time": "25.079",
"value": "first events were only open to Athenian males"
},
{
"operation": "Now on",
"time": "27.674",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "35.204",
"value": "Theseia"
},
{
"operation": "Now on",
"time": "36.047",
"value": "Theseia"
},
{
"operation": "Highlighting",
"time": "45.52",
"value": "Theseia_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "47.843",
"value": "Theseia_sentence_3"
},
{
"operation": "Highlighting",
"time": "49.137",
"value": "Theseia_sentence_4"
},
{
"operation": "Highlighting",
"time": "56.903",
"value": "Theseia_sentence_5"
},
{
"operation": "finish",
"time": "64.773",
"value": "finish"
}
]
}
|
14979
|
The Editio Regia which Estienne, that it was edited by in 1550 at Paris worked on one of the most important printed editions of the Greek New Testament in history making it the first Greek Testament that has a critical apparatus, an organized system of notations to represent, in a single text, the complex history of that text in a concise form, which is one of the most important printed editions of the Greek New Testament in history.
|
{
"evidence": [
"Editio Regia_sentence_1",
"Editio Regia_sentence_3",
"Editio Regia_sentence_4",
"Critical apparatus_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:\nThe Editio Regia which Estienne, that it was edited by in 1550 at Paris worked on one of the most important printed editions of the Greek New Testament in history making it the first Greek Testament that has a critical apparatus, an organized system of notations to represent, in a single text, the complex history of that text in a concise form, which is one of the most important printed editions of the Greek New Testament in history.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Editio Regia_sentence_1",
"Editio Regia_sentence_3",
"Editio Regia_sentence_4",
"Critical apparatus_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5865,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.549",
"value": "?search="
},
{
"operation": "search",
"time": "28.264",
"value": "Editio Regia"
},
{
"operation": "Now on",
"time": "28.969",
"value": "Editio Regia"
},
{
"operation": "Highlighting",
"time": "51.216",
"value": "Editio Regia_sentence_3"
},
{
"operation": "Highlighting",
"time": "77.248",
"value": "Editio Regia_sentence_1"
},
{
"operation": "Highlighting",
"time": "79",
"value": "Editio Regia_sentence_4"
},
{
"operation": "hyperlink",
"time": "110.862",
"value": "critical apparatus"
},
{
"operation": "Now on",
"time": "111.39",
"value": "Critical apparatus"
},
{
"operation": "Highlighting",
"time": "125.378",
"value": "Critical apparatus_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "129.111",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "129.241",
"value": "Editio Regia"
},
{
"operation": "Highlighting",
"time": "140.518",
"value": "Editio Regia_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "157.518",
"value": "Editio Regia_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "159.16",
"value": "Editio Regia_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "159.949",
"value": "Editio Regia_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "178.401",
"value": "Editio Regia_sentence_0"
},
{
"operation": "Highlighting",
"time": "180.107",
"value": "Editio Regia_sentence_1"
},
{
"operation": "Highlighting",
"time": "218.057",
"value": "Editio Regia_sentence_3"
},
{
"operation": "Highlighting",
"time": "231.389",
"value": "Editio Regia_sentence_4"
},
{
"operation": "hyperlink",
"time": "257.758",
"value": "critical apparatus"
},
{
"operation": "Now on",
"time": "258.505",
"value": "Critical apparatus"
},
{
"operation": "Highlighting",
"time": "262.047",
"value": "Critical apparatus_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "265.955",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "266.077",
"value": "Editio Regia"
},
{
"operation": "finish",
"time": "335.892",
"value": "finish"
}
]
}
|
36788
|
Beatrice Trew never taught at Lemsford, but she did meet her husband, a farmer named J. Albert Trew, there.
|
{
"evidence": [
"Beatrice Trew_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:\nBeatrice Trew never taught at Lemsford, but she did meet her husband, a farmer named J. Albert Trew, there.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Beatrice Trew_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5866,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.985",
"value": "?search="
},
{
"operation": "search",
"time": "9.989",
"value": "Beatrice Trew"
},
{
"operation": "Now on",
"time": "10.899",
"value": "Beatrice Trew"
},
{
"operation": "Page search",
"time": "23.027",
"value": "Lemsford"
},
{
"operation": "Highlighting",
"time": "38.124",
"value": "Beatrice Trew_sentence_2"
},
{
"operation": "finish",
"time": "52.521",
"value": "finish"
}
]
}
|
37665
|
The Tridrepana mediata is a butterfly species found in New Guinea.
|
{
"evidence": [
"Moth_sentence_1",
"Tridrepana mediata_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:\nThe Tridrepana mediata is a butterfly species found in New Guinea.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Moth_sentence_1",
"Tridrepana mediata_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5867,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.513",
"value": "undefined"
},
{
"operation": "search",
"time": "17.628",
"value": "Tridrepana mediata"
},
{
"operation": "Now on",
"time": "18.369",
"value": "Tridrepana mediata"
},
{
"operation": "Highlighting",
"time": "48.224",
"value": "Tridrepana mediata_sentence_2"
},
{
"operation": "hyperlink",
"time": "68.198",
"value": "moth"
},
{
"operation": "Now on",
"time": "68.822",
"value": "Moth"
},
{
"operation": "Highlighting",
"time": "80.038",
"value": "Moth_sentence_1"
},
{
"operation": "back-button-clicked",
"time": "113.716",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "113.776",
"value": "Tridrepana mediata"
},
{
"operation": "Highlighting",
"time": "115.374",
"value": "Tridrepana mediata_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "120.141",
"value": "Tridrepana mediata_sentence_2"
},
{
"operation": "finish",
"time": "136.409",
"value": "finish"
}
]
}
|
90955
|
The Chiman Saparaiya minister had a portfolio in agriculture as well as energy, after the first sworn minister Vijay Rupani in 2016.
|
{
"evidence": [
"Chiman Saparaiya_sentence_0",
"Chiman Saparaiya_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:\nThe Chiman Saparaiya minister had a portfolio in agriculture as well as energy, after the first sworn minister Vijay Rupani in 2016.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Chiman Saparaiya_sentence_0",
"Chiman Saparaiya_sentence_1"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5868,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.609",
"value": "?search="
},
{
"operation": "search",
"time": "39.387",
"value": "The Chiman Saparaiya minister"
},
{
"operation": "Now on",
"time": "40.315",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "57.262",
"value": "Chiman Saparaiya"
},
{
"operation": "Now on",
"time": "58.363",
"value": "Chiman Saparaiya"
},
{
"operation": "Highlighting",
"time": "68.678",
"value": "Chiman Saparaiya_sentence_0"
},
{
"operation": "Highlighting",
"time": "71.358",
"value": "Chiman Saparaiya_sentence_1"
},
{
"operation": "finish",
"time": "164.863",
"value": "finish"
}
]
}
|
80970
|
The orchestra, founded in 1966, gave their first subscription concert in May 1967; then in 1971, Hiroyuki Iwaki became the orchestra's first general music director, with the parallel appointment of Yoshikazu Fukumura as 'permanent conductor'.
|
{
"evidence": [
"Nagoya Philharmonic Orchestra_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:\nThe orchestra, founded in 1966, gave their first subscription concert in May 1967; then in 1971, Hiroyuki Iwaki became the orchestra's first general music director, with the parallel appointment of Yoshikazu Fukumura as 'permanent conductor'.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Nagoya Philharmonic Orchestra_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5869,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.758",
"value": "?search="
},
{
"operation": "search",
"time": "37.992",
"value": "Contains...Hiroyuki Iwaki became the orchestra's first general music director"
},
{
"operation": "Now on",
"time": "40.236",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "43.641",
"value": "Nagoya Philharmonic Orchestra"
},
{
"operation": "Now on",
"time": "44.039",
"value": "Nagoya Philharmonic Orchestra"
},
{
"operation": "Highlighting",
"time": "57.712",
"value": "Nagoya Philharmonic Orchestra_sentence_0"
},
{
"operation": "Highlighting",
"time": "62.044",
"value": "Nagoya Philharmonic Orchestra_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "63.419",
"value": "Nagoya Philharmonic Orchestra_sentence_0"
},
{
"operation": "finish",
"time": "78.679",
"value": "finish"
}
]
}
|
exfever_2487
|
Dennis Hopper was an American actor and filmmaker who appeared in the 1955 American epic Western drama film, directed by George Stevens.
|
{
"evidence": [
"Dennis Hopper",
"Giant (1956 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:\nDennis Hopper was an American actor and filmmaker who appeared in the 1955 American epic Western drama film, directed by George Stevens.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Dennis Hopper",
"Giant (1956 film)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5870,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6843
|
Siobhan Dowd died from cancer on August 21st at the early age of 47 will have shocked and saddened many PEN members.
|
{
"evidence": [
"A Monster Calls (film)",
"Patrick Ness"
],
"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:\nSiobhan Dowd died from cancer on August 21st at the early age of 47 will have shocked and saddened many PEN members.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"A Monster Calls (film)",
"Patrick Ness"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5871,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
51188
|
The Tunxis were a group of Connecticut Native Americans, their location makes them one of the Eastern Algonquian-speaking peoples and have never faced a disease outbreak.
|
{
"evidence": [
"Tunxis_sentence_2",
"Tunxis_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:\nThe Tunxis were a group of Connecticut Native Americans, their location makes them one of the Eastern Algonquian-speaking peoples and have never faced a disease outbreak.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Tunxis_sentence_2",
"Tunxis_sentence_5"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5872,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.433",
"value": "?search="
},
{
"operation": "search",
"time": "10.19",
"value": "Tunxis"
},
{
"operation": "Now on",
"time": "11.024",
"value": "Tunxis"
},
{
"operation": "Highlighting",
"time": "12.866",
"value": "Tunxis_sentence_2"
},
{
"operation": "Highlighting",
"time": "23.781",
"value": "Tunxis_sentence_3"
},
{
"operation": "Highlighting",
"time": "34.006",
"value": "Tunxis_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "39.307",
"value": "Tunxis_sentence_3"
},
{
"operation": "finish",
"time": "44.584",
"value": "finish"
}
]
}
|
90991
|
Nurfiled Priory is a hotel and health spa and is listed as a country house.
|
{
"evidence": [
"Nutfield Priory_sentence_2",
"Nutfield Priory_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:\nNurfiled Priory is a hotel and health spa and is listed as a country house.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Nutfield Priory_sentence_2",
"Nutfield Priory_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5873,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.507",
"value": "?search="
},
{
"operation": "search",
"time": "10.267",
"value": "Contains...Nurfiled Priory"
},
{
"operation": "Now on",
"time": "11.897",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "16.135",
"value": "List of monastic houses in Surrey"
},
{
"operation": "Now on",
"time": "17.451",
"value": "List of monastic houses in Surrey"
},
{
"operation": "search",
"time": "49.543",
"value": "Contains...Nurfield Priory"
},
{
"operation": "Now on",
"time": "51.112",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "55.664",
"value": "Nutfield Priory"
},
{
"operation": "Now on",
"time": "56.609",
"value": "Nutfield Priory"
},
{
"operation": "Highlighting",
"time": "60.226",
"value": "Nutfield Priory_sentence_2"
},
{
"operation": "Highlighting",
"time": "69.437",
"value": "Nutfield Priory_sentence_0"
},
{
"operation": "finish",
"time": "82.949",
"value": "finish"
}
]
}
|
61439
|
The idea of setting up an institution of the level of DMS was conceived by Mr. Shehzad Hassan Pervez, proposing that the medium of instruction was to be English and its students were to be prepared for Cambridge and BISE Faisalabad examinations.
|
{
"evidence": [
"Divisional Model College_sentence_8",
"Divisional Model College_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:\nThe idea of setting up an institution of the level of DMS was conceived by Mr. Shehzad Hassan Pervez, proposing that the medium of instruction was to be English and its students were to be prepared for Cambridge and BISE Faisalabad examinations.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Divisional Model College_sentence_8",
"Divisional Model College_sentence_9"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5874,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.471",
"value": "?search="
},
{
"operation": "search",
"time": "103.389",
"value": "Divisional Model College"
},
{
"operation": "Now on",
"time": "104.329",
"value": "Divisional Model College"
},
{
"operation": "Highlighting",
"time": "113.778",
"value": "Divisional Model College_sentence_8"
},
{
"operation": "Highlighting",
"time": "126.384",
"value": "Divisional Model College_sentence_9"
},
{
"operation": "finish",
"time": "134.402",
"value": "finish"
}
]
}
|
21395
|
RSI Corporation began in 1995 offering financial services.
|
{
"evidence": [
"RSI Corporation_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:\nRSI Corporation began in 1995 offering financial services.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"RSI Corporation_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5875,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.093",
"value": "?search="
},
{
"operation": "search",
"time": "112.029",
"value": "RSI Corporation"
},
{
"operation": "Now on",
"time": "112.935",
"value": "RSI Corporation"
},
{
"operation": "Page search",
"time": "149.801",
"value": "financial"
},
{
"operation": "Page search",
"time": "160.55",
"value": "1995"
},
{
"operation": "Highlighting",
"time": "179.662",
"value": "RSI Corporation_sentence_2"
},
{
"operation": "finish",
"time": "195.735",
"value": "finish"
}
]
}
|
76659
|
Murray Aynsley Hill, a suburb of Christchurch, New Zealand was named after early settler Hugh Murray-Aynsley who was also the first owner of the land.
|
{
"evidence": [
"Murray Aynsley Hill_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:\nMurray Aynsley Hill, a suburb of Christchurch, New Zealand was named after early settler Hugh Murray-Aynsley who was also the first owner of the land.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Murray Aynsley Hill_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5876,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.914",
"value": "undefined"
},
{
"operation": "search",
"time": "5.925",
"value": "Murray Aynsley Hill"
},
{
"operation": "Now on",
"time": "7.203",
"value": "Murray Aynsley Hill"
},
{
"operation": "Highlighting",
"time": "9.767",
"value": "Murray Aynsley Hill_sentence_0"
},
{
"operation": "Highlighting",
"time": "14.71",
"value": "Murray Aynsley Hill_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "18.458",
"value": "Murray Aynsley Hill_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "19.395",
"value": "Murray Aynsley Hill_sentence_0"
},
{
"operation": "Highlighting",
"time": "20.443",
"value": "Murray Aynsley Hill_sentence_3"
},
{
"operation": "finish",
"time": "29.988",
"value": "finish"
}
]
}
|
56590
|
Konstantin Dumba had admitted he'd given a letter to James Francis Jewell Archibald, a man who would later be arrested after WWI.
|
{
"evidence": [
"Konstantin Dumba_sentence_15",
"James Francis Jewell Archibald_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:\nKonstantin Dumba had admitted he'd given a letter to James Francis Jewell Archibald, a man who would later be arrested after WWI.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Konstantin Dumba_sentence_15",
"James Francis Jewell Archibald_sentence_2"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5877,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.092",
"value": "?search="
},
{
"operation": "search",
"time": "11.437",
"value": "Konstantin Dumba"
},
{
"operation": "Now on",
"time": "13.242",
"value": "Konstantin Dumba"
},
{
"operation": "Page search",
"time": "16.254",
"value": "archi"
},
{
"operation": "Highlighting",
"time": "31.71",
"value": "Konstantin Dumba_sentence_15"
},
{
"operation": "hyperlink",
"time": "37.745",
"value": "James Francis Jewell Archibald"
},
{
"operation": "Now on",
"time": "38.289",
"value": "James Francis Jewell Archibald"
},
{
"operation": "Highlighting",
"time": "42.048",
"value": "James Francis Jewell Archibald_sentence_2"
},
{
"operation": "finish",
"time": "50.918",
"value": "finish"
}
]
}
|
exfever_3119
|
The X Factor is an American television music competition to find new singing talent, presented by Konnie Huq on ITV2, who took over from an English television presenter, model, and author.
|
{
"evidence": [
"The X Factor (UK series 7)",
"Holly Willoughby"
],
"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 X Factor is an American television music competition to find new singing talent, presented by Konnie Huq on ITV2, who took over from an English television presenter, model, and author.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The X Factor (UK series 7)",
"Holly Willoughby"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5878,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
68094
|
In the 2017-2018 season, Albania Erblira Bici was number one as an opposite.
|
{
"evidence": [
"Erblira Bici_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:\nIn the 2017-2018 season, Albania Erblira Bici was number one as an opposite.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Erblira Bici_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 5879,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "4.452",
"value": "?search="
},
{
"operation": "search",
"time": "59.266",
"value": "Albania Erblira Bici"
},
{
"operation": "Now on",
"time": "60.002",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "62.55",
"value": "Erblira Bici"
},
{
"operation": "Now on",
"time": "63.076",
"value": "Erblira Bici"
},
{
"operation": "Highlighting",
"time": "164.408",
"value": "Erblira Bici_sentence_0"
},
{
"operation": "finish",
"time": "288.053",
"value": "finish"
}
]
}
|
67572
|
Blue Nile patas monkey, which was first described in 1962 and then reclassified in 2010, has a distinctive hair on its face that resembles a Walrus moustache.
|
{
"evidence": [
"Blue Nile patas monkey_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:\nBlue Nile patas monkey, which was first described in 1962 and then reclassified in 2010, has a distinctive hair on its face that resembles a Walrus moustache.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Blue Nile patas monkey_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 5880,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.821",
"value": "?search="
},
{
"operation": "search",
"time": "337.752",
"value": "Blue Nile patas monkey"
},
{
"operation": "Now on",
"time": "339.253",
"value": "Blue Nile patas monkey"
},
{
"operation": "Highlighting",
"time": "399.941",
"value": "Blue Nile patas monkey_sentence_2"
},
{
"operation": "finish",
"time": "410.044",
"value": "finish"
}
]
}
|
exfever_5070
|
Diane Ruggiero is a television producer who writes a drama series produced by a television filmmaker.
|
{
"evidence": [
"Diane Ruggiero",
"Dirty Sexy Money",
"Greg Berlanti"
],
"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:\nDiane Ruggiero is a television producer who writes a drama series produced by a television filmmaker.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Diane Ruggiero",
"Dirty Sexy Money",
"Greg Berlanti"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5881,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8995
|
Albania is a country in Southeastern Europe that shares land borders with a partially recognized state that declared its independence from a landlocked country in Southeast Europe.
|
{
"evidence": [
"Albania",
"Kosovo",
"Serbia"
],
"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:\nAlbania is a country in Southeastern Europe that shares land borders with a partially recognized state that declared its independence from a landlocked country in Southeast Europe.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Albania",
"Kosovo",
"Serbia"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5882,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
60939
|
SV Lurup is a German association football club from the city of Hamburg in the federal state of the same name, The club was founded in 1923 as Turn- und Sportverein Lurup in the former town Lurup and later that same year merged with FSV 1920 Schenefeld to form TSV Lurup-Schenefeld in a union that lasted until 1930 when the two clubs went their separate ways.
|
{
"evidence": [
"SV Lurup_sentence_0",
"SV Lurup_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:\nSV Lurup is a German association football club from the city of Hamburg in the federal state of the same name, The club was founded in 1923 as Turn- und Sportverein Lurup in the former town Lurup and later that same year merged with FSV 1920 Schenefeld to form TSV Lurup-Schenefeld in a union that lasted until 1930 when the two clubs went their separate ways.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"SV Lurup_sentence_0",
"SV Lurup_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 5883,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.743",
"value": "?search="
},
{
"operation": "search",
"time": "12.717",
"value": "SV Lurup"
},
{
"operation": "Now on",
"time": "13.414",
"value": "SV Lurup"
},
{
"operation": "Highlighting",
"time": "20.184",
"value": "SV Lurup_sentence_0"
},
{
"operation": "Highlighting",
"time": "49.945",
"value": "SV Lurup_sentence_1"
},
{
"operation": "finish",
"time": "70.628",
"value": "finish"
}
]
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.