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_3606
|
J. Marshall II lived in America with his family.
|
{
"evidence": [
"J. Howard Marshall",
"Anna Nicole Smith"
],
"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:\nJ. Marshall II lived in America with his family.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"J. Howard Marshall",
"Anna Nicole Smith"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4484,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
52723
|
The state of Victoria in Australia, the second-smallest state with a land area of 91,761 square miles, experienced the wettest January in 2011 which caused flash flooding, resulting in damage to businesses and homes and swelling in the waterways.
|
{
"evidence": [
"Victoria (Australia)_sentence_2",
"Victoria (Australia)_sentence_3",
"Victoria (Australia)_sentence_123",
"Wilsons Promontory_sentence_31"
],
"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 state of Victoria in Australia, the second-smallest state with a land area of 91,761 square miles, experienced the wettest January in 2011 which caused flash flooding, resulting in damage to businesses and homes and swelling in the waterways.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Victoria (Australia)_sentence_2",
"Victoria (Australia)_sentence_3",
"Victoria (Australia)_sentence_123",
"Wilsons Promontory_sentence_31"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4485,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.067",
"value": "?search="
},
{
"operation": "search",
"time": "18.978",
"value": "Victoria, Australia"
},
{
"operation": "Now on",
"time": "28.768",
"value": "Victoria (Australia)"
},
{
"operation": "Highlighting",
"time": "33.421",
"value": "Victoria (Australia)_sentence_2"
},
{
"operation": "Highlighting",
"time": "37.9",
"value": "Victoria (Australia)_sentence_3"
},
{
"operation": "Page search",
"time": "52.994",
"value": "2011"
},
{
"operation": "Highlighting",
"time": "77.069",
"value": "Victoria (Australia)_sentence_123"
},
{
"operation": "hyperlink",
"time": "141.451",
"value": "Tidal River"
},
{
"operation": "Now on",
"time": "142.091",
"value": "Tidal River, Victoria"
},
{
"operation": "Page search",
"time": "149.232",
"value": "2011"
},
{
"operation": "hyperlink",
"time": "160.521",
"value": "Tidal River (Victoria)"
},
{
"operation": "Now on",
"time": "161.437",
"value": "Tidal River (Victoria)"
},
{
"operation": "Page search",
"time": "164.689",
"value": "2011"
},
{
"operation": "search",
"time": "186.577",
"value": "victoria, australia flooding"
},
{
"operation": "Now on",
"time": "187.823",
"value": "undefined"
},
{
"operation": "search",
"time": "202.442",
"value": "2011 flood victoria australia"
},
{
"operation": "Now on",
"time": "203.163",
"value": "undefined"
},
{
"operation": "search",
"time": "214.531",
"value": "Victoria, Australia"
},
{
"operation": "Now on",
"time": "215.108",
"value": "Victoria, Australia"
},
{
"operation": "hyperlink",
"time": "227.094",
"value": "2.2.1 Rainfall Victoria (Australia)_section_8"
},
{
"operation": "hyperlink",
"time": "231.068",
"value": "Wilsons Promontory National Park"
},
{
"operation": "Now on",
"time": "231.605",
"value": "Wilsons Promontory National Park"
},
{
"operation": "hyperlink",
"time": "254.56",
"value": "Wilsons Promontory"
},
{
"operation": "Now on",
"time": "256.544",
"value": "Wilsons Promontory"
},
{
"operation": "Highlighting",
"time": "260.659",
"value": "Wilsons Promontory_sentence_31"
},
{
"operation": "hyperlink",
"time": "346.257",
"value": "Victoria"
},
{
"operation": "Now on",
"time": "347.808",
"value": "Victoria (Australia)"
},
{
"operation": "finish",
"time": "392.464",
"value": "finish"
}
]
}
|
42271
|
Marilyn Miller, who died following nasal surgery, performed in Ziegfeld Follies with Florenz Ziegfeld in 1981.
|
{
"evidence": [
"Marilyn Miller_sentence_10"
],
"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:\nMarilyn Miller, who died following nasal surgery, performed in Ziegfeld Follies with Florenz Ziegfeld in 1981.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Marilyn Miller_sentence_10"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4486,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.233",
"value": "?search="
},
{
"operation": "search",
"time": "10.772",
"value": "Marilyn Miller"
},
{
"operation": "Now on",
"time": "11.71",
"value": "Marilyn Miller"
},
{
"operation": "Highlighting",
"time": "54.749",
"value": "Marilyn Miller_sentence_25"
},
{
"operation": "Highlighting",
"time": "70.484",
"value": "Marilyn Miller_sentence_26"
},
{
"operation": "Highlighting",
"time": "73.721",
"value": "Marilyn Miller_sentence_27"
},
{
"operation": "Highlighting",
"time": "108.252",
"value": "Marilyn Miller_sentence_37"
},
{
"operation": "Highlighting",
"time": "108.257",
"value": "Marilyn Miller_section_3"
},
{
"operation": "Highlighting",
"time": "108.259",
"value": "Marilyn Miller_section_5"
},
{
"operation": "Highlighting",
"time": "121.574",
"value": "Marilyn Miller_sentence_35"
},
{
"operation": "Highlighting deleted",
"time": "123.044",
"value": "Marilyn Miller_sentence_37"
},
{
"operation": "Highlighting",
"time": "123.7",
"value": "Marilyn Miller_sentence_36"
},
{
"operation": "Highlighting",
"time": "124.141",
"value": "Marilyn Miller_sentence_37"
},
{
"operation": "Highlighting deleted",
"time": "155.626",
"value": "Marilyn Miller_sentence_25"
},
{
"operation": "Highlighting deleted",
"time": "156.227",
"value": "Marilyn Miller_sentence_26"
},
{
"operation": "Highlighting deleted",
"time": "156.703",
"value": "Marilyn Miller_sentence_27"
},
{
"operation": "Page search",
"time": "160.963",
"value": "foll"
},
{
"operation": "Highlighting",
"time": "172.512",
"value": "Marilyn Miller_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "175.017",
"value": "Marilyn Miller_header_cell_0_3_0"
},
{
"operation": "Highlighting",
"time": "175.834",
"value": "Marilyn Miller_cell_0_3_1"
},
{
"operation": "Highlighting",
"time": "175.872",
"value": "Marilyn Miller_header_cell_0_3_0"
},
{
"operation": "Highlighting",
"time": "175.884",
"value": "Marilyn Miller_header_cell_0_0_0"
},
{
"operation": "Highlighting deleted",
"time": "186.656",
"value": "Marilyn Miller_sentence_37"
},
{
"operation": "Highlighting deleted",
"time": "187.261",
"value": "Marilyn Miller_sentence_36"
},
{
"operation": "Highlighting deleted",
"time": "187.734",
"value": "Marilyn Miller_sentence_35"
},
{
"operation": "Highlighting deleted",
"time": "187.741",
"value": "Marilyn Miller_section_3"
},
{
"operation": "Highlighting deleted",
"time": "187.747",
"value": "Marilyn Miller_section_5"
},
{
"operation": "Highlighting",
"time": "207.84",
"value": "Marilyn Miller_sentence_10"
},
{
"operation": "Highlighting",
"time": "210.08",
"value": "Marilyn Miller_sentence_13"
},
{
"operation": "Highlighting deleted",
"time": "230.346",
"value": "Marilyn Miller_sentence_13"
},
{
"operation": "Highlighting deleted",
"time": "237.135",
"value": "Marilyn Miller_cell_0_3_1"
},
{
"operation": "Highlighting deleted",
"time": "237.154",
"value": "Marilyn Miller_header_cell_0_3_0"
},
{
"operation": "Highlighting deleted",
"time": "237.158",
"value": "Marilyn Miller_header_cell_0_0_0"
},
{
"operation": "finish",
"time": "250.041",
"value": "finish"
}
]
}
|
exfever_3027
|
Coldplay is a British rock band whose members include an English singer-songwriter who was a co-founder of the rock band Linkin Park.
|
{
"evidence": [
"Coldplay",
"Chris Martin"
],
"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:\nColdplay is a British rock band whose members include an English singer-songwriter who was a co-founder of the rock band Linkin Park.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Coldplay",
"Chris Martin"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4487,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
51346
|
Judith Pella earned a Bachelor of Arts degree in Social Science, she worked as an instructional teacher’s aide and most of her books never needed research.
|
{
"evidence": [
"Judith Pella_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:\nJudith Pella earned a Bachelor of Arts degree in Social Science, she worked as an instructional teacher’s aide and most of her books never needed research.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Judith Pella_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4488,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.077",
"value": "?search="
},
{
"operation": "search",
"time": "21.276",
"value": "Judith Pella"
},
{
"operation": "Now on",
"time": "22.027",
"value": "Judith Pella"
},
{
"operation": "Highlighting",
"time": "28.102",
"value": "Judith Pella_sentence_2"
},
{
"operation": "Highlighting",
"time": "28.103",
"value": "Judith Pella_section_0"
},
{
"operation": "Highlighting",
"time": "32.738",
"value": "Judith Pella_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "38.059",
"value": "Judith Pella_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "38.474",
"value": "Judith Pella_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "38.477",
"value": "Judith Pella_section_0"
},
{
"operation": "Highlighting",
"time": "39.217",
"value": "Judith Pella_sentence_1"
},
{
"operation": "finish",
"time": "46.181",
"value": "finish"
}
]
}
|
52249
|
On 11 September 1907, Grumel entered the Assumptionist novitiate of Louvain, Belgium (25 kilometres east of Brussels); then took philosophy courses in Rome where he obtained a degree from 1909 to 1911.
|
{
"evidence": [
"Venance Grumel_sentence_6",
"Venance Grumel_sentence_8",
"Leuven_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:\nOn 11 September 1907, Grumel entered the Assumptionist novitiate of Louvain, Belgium (25 kilometres east of Brussels); then took philosophy courses in Rome where he obtained a degree from 1909 to 1911.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Venance Grumel_sentence_6",
"Venance Grumel_sentence_8",
"Leuven_sentence_3"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4489,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.969",
"value": "?search="
},
{
"operation": "search",
"time": "24.144",
"value": "Grumel"
},
{
"operation": "Now on",
"time": "24.826",
"value": "Venance Grumel"
},
{
"operation": "Highlighting",
"time": "29.335",
"value": "Venance Grumel_sentence_6"
},
{
"operation": "Highlighting",
"time": "42.711",
"value": "Venance Grumel_sentence_8"
},
{
"operation": "hyperlink",
"time": "47.826",
"value": "Louvain, Belgium"
},
{
"operation": "Now on",
"time": "49.071",
"value": "Leuven"
},
{
"operation": "Highlighting",
"time": "56.356",
"value": "Leuven_sentence_3"
},
{
"operation": "finish",
"time": "185.787",
"value": "finish"
}
]
}
|
52250
|
Political scientist Charles Murray gave A Troublesome Inheritance a more favorable review than others.
|
{
"evidence": [
"A Troublesome Inheritance_sentence_20"
],
"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:\nPolitical scientist Charles Murray gave A Troublesome Inheritance a more favorable review than others.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"A Troublesome Inheritance_sentence_20"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4490,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "14.909",
"value": "?search="
},
{
"operation": "search",
"time": "55.666",
"value": "Charles Murray"
},
{
"operation": "Now on",
"time": "56.575",
"value": "Charles Murray"
},
{
"operation": "search",
"time": "78.722",
"value": "A Troublesome Inheritance"
},
{
"operation": "Now on",
"time": "79.47",
"value": "A Troublesome Inheritance"
},
{
"operation": "Page search",
"time": "87.107",
"value": "charles"
},
{
"operation": "Highlighting",
"time": "116.534",
"value": "A Troublesome Inheritance_sentence_20"
},
{
"operation": "finish",
"time": "135.689",
"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": 4491,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
88318
|
Driven by Takuma Sato, the Super Aguri SA06 was the car with which the Super Aguri team competed in the latter part of the 2006 Formula One season, the 60th season of Formula One motor racing.
|
{
"evidence": [
"Super Aguri SA06_sentence_0",
"Super Aguri SA06_sentence_1",
"2006 Formula One World Championship_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:\nDriven by Takuma Sato, the Super Aguri SA06 was the car with which the Super Aguri team competed in the latter part of the 2006 Formula One season, the 60th season of Formula One motor racing.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Super Aguri SA06_sentence_0",
"Super Aguri SA06_sentence_1",
"2006 Formula One World Championship_sentence_2"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4492,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.966",
"value": "?search="
},
{
"operation": "search",
"time": "35.932",
"value": "Super Aguri SA06"
},
{
"operation": "Now on",
"time": "37.031",
"value": "Super Aguri SA06"
},
{
"operation": "Highlighting",
"time": "49.498",
"value": "Super Aguri SA06_sentence_0"
},
{
"operation": "Highlighting",
"time": "56.11",
"value": "Super Aguri SA06_sentence_1"
},
{
"operation": "hyperlink",
"time": "62.637",
"value": "2006"
},
{
"operation": "Now on",
"time": "64.211",
"value": "2006 Formula One World Championship"
},
{
"operation": "Highlighting",
"time": "69.306",
"value": "2006 Formula One World Championship_sentence_2"
},
{
"operation": "finish",
"time": "78.881",
"value": "finish"
}
]
}
|
exfever_6149
|
Kirk Douglas made his film debut in The Strange Love of Martha Ivers with an American actress known for her strong, realistic screen presence and versatility.
|
{
"evidence": [
"Kirk Douglas",
"Barbara Stanwyck"
],
"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:\nKirk Douglas made his film debut in The Strange Love of Martha Ivers with an American actress known for her strong, realistic screen presence and versatility.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Kirk Douglas",
"Barbara Stanwyck"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4493,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_7424
|
Odia is a branch of the Indo-Iranian languages spoken natively by the Indo-Aryan peoples.
|
{
"evidence": [
"Odia language",
"Indo-Aryan languages"
],
"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:\nOdia is a branch of the Indo-Iranian languages spoken natively by the Indo-Aryan peoples.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Odia language",
"Indo-Aryan languages"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4494,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
6437
|
Vratsa waterfalls(Skaklia, 141 metres (462 feet)and Borov Kamak, 63 metres) are waterfalls that overlook the town of Vratsa and the Vratsa lowlands and where remains from the ancient Bulgarian village Patleina were found
|
{
"evidence": [
"Vratsa waterfalls_sentence_0",
"Vratsa waterfalls_sentence_1",
"Vratsa waterfalls_sentence_3",
"Vratsa waterfalls_sentence_2",
"Vratsa waterfalls_sentence_8",
"Vratsa waterfalls_sentence_9",
"Vratsa waterfalls_sentence_4",
"Vratsa waterfalls_sentence_5"
],
"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:\nVratsa waterfalls(Skaklia, 141 metres (462 feet)and Borov Kamak, 63 metres) are waterfalls that overlook the town of Vratsa and the Vratsa lowlands and where remains from the ancient Bulgarian village Patleina were found\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Vratsa waterfalls_sentence_0",
"Vratsa waterfalls_sentence_1",
"Vratsa waterfalls_sentence_3",
"Vratsa waterfalls_sentence_2",
"Vratsa waterfalls_sentence_8",
"Vratsa waterfalls_sentence_9",
"Vratsa waterfalls_sentence_4",
"Vratsa waterfalls_sentence_5"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4495,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.891",
"value": "?search="
},
{
"operation": "search",
"time": "19.271",
"value": "Vratsa waterfalls"
},
{
"operation": "Now on",
"time": "20.742",
"value": "Vratsa waterfalls"
},
{
"operation": "Highlighting",
"time": "35.204",
"value": "Vratsa waterfalls_sentence_0"
},
{
"operation": "Highlighting",
"time": "38.211",
"value": "Vratsa waterfalls_sentence_1"
},
{
"operation": "Highlighting",
"time": "67.3",
"value": "Vratsa waterfalls_sentence_3"
},
{
"operation": "Highlighting",
"time": "73.715",
"value": "Vratsa waterfalls_sentence_2"
},
{
"operation": "Highlighting",
"time": "119.562",
"value": "Vratsa waterfalls_sentence_8"
},
{
"operation": "Page search",
"time": "178.607",
"value": "lowlands"
},
{
"operation": "Highlighting",
"time": "185.571",
"value": "Vratsa waterfalls_sentence_9"
},
{
"operation": "Highlighting",
"time": "216.561",
"value": "Vratsa waterfalls_sentence_4"
},
{
"operation": "Highlighting",
"time": "217.268",
"value": "Vratsa waterfalls_sentence_5"
},
{
"operation": "finish",
"time": "242.004",
"value": "finish"
}
]
}
|
16205
|
Zach Collaros played with the Cincinnati Bearcats from 2008 - 2011, which Brian Kelly then coached until 2009, and by Bruce Arians from 2010 - 2011.
|
{
"evidence": [
"Cincinnati Bearcats football_sentence_61",
"Cincinnati Bearcats football_sentence_80"
],
"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:\nZach Collaros played with the Cincinnati Bearcats from 2008 - 2011, which Brian Kelly then coached until 2009, and by Bruce Arians from 2010 - 2011.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Cincinnati Bearcats football_sentence_61",
"Cincinnati Bearcats football_sentence_80"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4496,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.001",
"value": "?search="
},
{
"operation": "search",
"time": "13.241",
"value": "Zach Collaros"
},
{
"operation": "Now on",
"time": "14.191",
"value": "Zach Collaros"
},
{
"operation": "Highlighting",
"time": "59.428",
"value": "Zach Collaros_sentence_5"
},
{
"operation": "Highlighting",
"time": "59.431",
"value": "Zach Collaros_title"
},
{
"operation": "Highlighting",
"time": "59.432",
"value": "Zach Collaros_section_1"
},
{
"operation": "Highlighting",
"time": "73.405",
"value": "Zach Collaros_sentence_16"
},
{
"operation": "Highlighting",
"time": "97.243",
"value": "Zach Collaros_sentence_3"
},
{
"operation": "search",
"time": "122.499",
"value": "Cincinnati Bearcats football"
},
{
"operation": "Now on",
"time": "124.017",
"value": "Cincinnati Bearcats football"
},
{
"operation": "hyperlink",
"time": "139.956",
"value": "1.13 Brian Kelly era (2007–2009) Cincinnati Bearcats football_section_13"
},
{
"operation": "Highlighting",
"time": "148.138",
"value": "Cincinnati Bearcats football_sentence_61"
},
{
"operation": "Highlighting",
"time": "173.429",
"value": "Cincinnati Bearcats football_sentence_80"
},
{
"operation": "finish",
"time": "247.377",
"value": "finish"
}
]
}
|
exfever_7598
|
Willie Hugh Nelson was the first artist recognized in country music.
|
{
"evidence": [
"Willie Nelson",
"Shotgun Willie"
],
"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:\nWillie Hugh Nelson was the first artist recognized in country music.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Willie Nelson",
"Shotgun Willie"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4497,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6746
|
The Vampire Diaries is a supernatural teen drama TV series that premiered on an American English-language commercial broadcast television network.
|
{
"evidence": [
"The Vampire Diaries",
"The CW"
],
"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 Vampire Diaries is a supernatural teen drama TV series that premiered on an American English-language commercial broadcast television network.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Vampire Diaries",
"The CW"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4498,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8573
|
The Mummy is a 2017 American fantasy action-adventure film directed by Alex Kurtzman and written by an American screenwriter, film director, and producer who made his directorial debut with "Jack Reacher."
|
{
"evidence": [
"The Mummy (2017 film)",
"Christopher McQuarrie"
],
"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 Mummy is a 2017 American fantasy action-adventure film directed by Alex Kurtzman and written by an American screenwriter, film director, and producer who made his directorial debut with \"Jack Reacher.\"\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Mummy (2017 film)",
"Christopher McQuarrie"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4499,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6565
|
Margaret Constance is the first English actress who developed and co-launched the social media platform, Daisie.
|
{
"evidence": [
"Maisie Williams",
"Doctor Who"
],
"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:\nMargaret Constance is the first English actress who developed and co-launched the social media platform, Daisie.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Maisie Williams",
"Doctor Who"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4500,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5592
|
Lung cancer is a malignant neoplastic lung nodule.
|
{
"evidence": [
"Lung cancer",
"Lung tumor"
],
"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:\nLung cancer is a malignant neoplastic lung nodule.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Lung cancer",
"Lung tumor"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4501,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
26419
|
The teams that compete in Centrobasket are explicitly not part of the CONCENCABA, and the winners of the Centrobasket can proceed to the Summer Olympics or to the FIFA World Cup.
|
{
"evidence": [
"Centrobasket_sentence_0",
"Centrobasket_sentence_1"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe teams that compete in Centrobasket are explicitly not part of the CONCENCABA, and the winners of the Centrobasket can proceed to the Summer Olympics or to the FIFA World Cup.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Centrobasket_sentence_0",
"Centrobasket_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4502,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.693",
"value": "?search="
},
{
"operation": "search",
"time": "25.101",
"value": "Centrobasket"
},
{
"operation": "Now on",
"time": "29.657",
"value": "Centrobasket"
},
{
"operation": "Highlighting",
"time": "73.753",
"value": "Centrobasket_sentence_0"
},
{
"operation": "Highlighting",
"time": "74.688",
"value": "Centrobasket_sentence_1"
},
{
"operation": "finish",
"time": "118.68",
"value": "finish"
}
]
}
|
exfever_1476
|
Kathrin Romany Beckinsale is an actress and model who appeared in a 1999 American drama film directed by Jonathan Kaplan.
|
{
"evidence": [
"Kate Beckinsale",
"Brokedown Palace"
],
"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:\nKathrin Romany Beckinsale is an actress and model who appeared in a 1999 American drama film directed by Jonathan Kaplan.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Kate Beckinsale",
"Brokedown Palace"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4503,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
57984
|
Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg moved to Altenburg in Thuringia in 1604, releasing publications the entire time (though its most prominent one was in 1606).
|
{
"evidence": [
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_7",
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_0",
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_11",
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_8"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nFürstlich Sächsischer Hofbuchdruckerei zu Altenburg moved to Altenburg in Thuringia in 1604, releasing publications the entire time (though its most prominent one was in 1606).\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_7",
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_0",
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_11",
"Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_8"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4504,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.675",
"value": "?search="
},
{
"operation": "search",
"time": "8.058",
"value": "Fürstlich Sächsischer"
},
{
"operation": "Now on",
"time": "29.489",
"value": "?search="
},
{
"operation": "search",
"time": "31.792",
"value": "Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg"
},
{
"operation": "Now on",
"time": "33.382",
"value": "Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg"
},
{
"operation": "Highlighting",
"time": "82.011",
"value": "Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_7"
},
{
"operation": "Page search",
"time": "102.482",
"value": "Thuringia"
},
{
"operation": "Highlighting",
"time": "117.045",
"value": "Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_0"
},
{
"operation": "Highlighting",
"time": "138.235",
"value": "Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_11"
},
{
"operation": "Highlighting",
"time": "227.099",
"value": "Fürstlich Sächsischer Hofbuchdruckerei zu Altenburg_sentence_8"
},
{
"operation": "finish",
"time": "304.446",
"value": "finish"
}
]
}
|
15335
|
Kaneshiro Kōfuku was a professional surfer from Takanabe, Miyazaki, Japan and wrestled from 1969 to 1987.
|
{
"evidence": [
"Kaneshiro Kōfuku_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:\nKaneshiro Kōfuku was a professional surfer from Takanabe, Miyazaki, Japan and wrestled from 1969 to 1987.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Kaneshiro Kōfuku_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4505,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.603",
"value": "?search="
},
{
"operation": "search",
"time": "5.146",
"value": "Kaneshiro Kōfuku"
},
{
"operation": "Now on",
"time": "5.992",
"value": "Kaneshiro Kōfuku"
},
{
"operation": "Highlighting",
"time": "9.934",
"value": "Kaneshiro Kōfuku_sentence_1"
},
{
"operation": "finish",
"time": "19.523",
"value": "finish"
}
]
}
|
exfever_7988
|
Gerald Ford served as the leader of a contemporary political party in the United States that was simultaneously strengthened by the collapse of the political party that espoused traditionalist conservatism during the middle of the 19th century.
|
{
"evidence": [
"Gerald Ford",
"Republican Party (United States)",
"Whig Party (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:\nGerald Ford served as the leader of a contemporary political party in the United States that was simultaneously strengthened by the collapse of the political party that espoused traditionalist conservatism during the middle of the 19th century.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Gerald Ford",
"Republican Party (United States)",
"Whig Party (United States)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4506,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_963
|
Liverpool Football Club is a professional basketball club based in Liverpool, England, that has won seven titles in an annual knockout football competition in men's domestic English football.
|
{
"evidence": [
"Liverpool F.C.",
"FA Cup"
],
"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:\nLiverpool Football Club is a professional basketball club based in Liverpool, England, that has won seven titles in an annual knockout football competition in men's domestic English football.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Liverpool F.C.",
"FA Cup"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4507,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8201
|
The cell came from the dialect spoken in present-day Rome, then known as a region in central western Italy.
|
{
"evidence": [
"Cell (biology)",
"Latin",
"Latium"
],
"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 cell came from the dialect spoken in present-day Rome, then known as a region in central western Italy.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Cell (biology)",
"Latin",
"Latium"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4508,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
45393
|
Fahd bin Muqrin (son of former Crown Prince Muqrin bin Abdul-Aziz) is the president of Sons of the Homeland and Saudi National Environmental Education Initiative
|
{
"evidence": [
"Fahd bin Muqrin_sentence_1",
"Fahd bin Muqrin_sentence_2",
"Fahd bin Muqrin_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:\nFahd bin Muqrin (son of former Crown Prince Muqrin bin Abdul-Aziz) is the president of Sons of the Homeland and Saudi National Environmental Education Initiative\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Fahd bin Muqrin_sentence_1",
"Fahd bin Muqrin_sentence_2",
"Fahd bin Muqrin_sentence_3"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4509,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.91",
"value": "?search="
},
{
"operation": "search",
"time": "17.327",
"value": "Fahd bin Muqrin"
},
{
"operation": "Now on",
"time": "18.286",
"value": "Fahd bin Muqrin"
},
{
"operation": "Page search",
"time": "36.072",
"value": "Sons of the Homeland"
},
{
"operation": "Highlighting",
"time": "50.566",
"value": "Fahd bin Muqrin_sentence_1"
},
{
"operation": "Highlighting",
"time": "58.347",
"value": "Fahd bin Muqrin_sentence_2"
},
{
"operation": "Highlighting",
"time": "70.649",
"value": "Fahd bin Muqrin_sentence_3"
},
{
"operation": "finish",
"time": "85.631",
"value": "finish"
}
]
}
|
exfever_6131
|
Django Unchained is a 2017 American revisionist Western film written and directed by Quentin Tarantino and starring an American actor who is one of the most widely recognized actors of his generation.
|
{
"evidence": [
"Django Unchained",
"Samuel L. Jackson"
],
"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:\nDjango Unchained is a 2017 American revisionist Western film written and directed by Quentin Tarantino and starring an American actor who is one of the most widely recognized actors of his generation.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Django Unchained",
"Samuel L. Jackson"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4510,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
48974
|
Some of the principals that worked at Enchi College of Education (affiliated with Abetifi Presbyterian College of Education, the oldest and largest of the 13 Ghanaian public universities) are Mr. Djeo Addison (who served from 1965 until 1973), Mr. D.M. Korwumu (who served from 1974-1976), Mr. John Attipila (who served from 1976-1980), and Mr. Assifuah Nunoo (who served from 1980-1983).
|
{
"evidence": [
"Enchi College of Education_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:\nSome of the principals that worked at Enchi College of Education (affiliated with Abetifi Presbyterian College of Education, the oldest and largest of the 13 Ghanaian public universities) are Mr. Djeo Addison (who served from 1965 until 1973), Mr. D.M. Korwumu (who served from 1974-1976), Mr. John Attipila (who served from 1976-1980), and Mr. Assifuah Nunoo (who served from 1980-1983).\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Enchi College of Education_sentence_4"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4511,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.147",
"value": "?search="
},
{
"operation": "search",
"time": "17.419",
"value": "Enchi College of Education"
},
{
"operation": "Now on",
"time": "18.328",
"value": "Enchi College of Education"
},
{
"operation": "Page search",
"time": "35.896",
"value": "presbyterian"
},
{
"operation": "Page search",
"time": "62.232",
"value": "affili"
},
{
"operation": "Highlighting",
"time": "65.993",
"value": "Enchi College of Education_sentence_4"
},
{
"operation": "finish",
"time": "179.324",
"value": "finish"
}
]
}
|
exfever_8758
|
World War I was referred to by its contemporaries as the "Great War", in which belligerents included much of a continent located entirely in the Northern Hemisphere and mostly in the geographical term for the half of Earth which is east of the prime meridian and west of the antimeridian.
|
{
"evidence": [
"World War I",
"Europe",
"Eastern Hemisphere"
],
"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:\nWorld War I was referred to by its contemporaries as the \"Great War\", in which belligerents included much of a continent located entirely in the Northern Hemisphere and mostly in the geographical term for the half of Earth which is east of the prime meridian and west of the antimeridian.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"World War I",
"Europe",
"Eastern Hemisphere"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4512,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3697
|
The Adjustment Bureau is based on a short story by an American science fiction writer who explored varied social questions, including a concept that denotes the fundamental dispositions that humans are said to have naturally.?
|
{
"evidence": [
"The Adjustment Bureau",
"Philip K. Dick",
"Human nature"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Adjustment Bureau is based on a short story by an American science fiction writer who explored varied social questions, including a concept that denotes the fundamental dispositions that humans are said to have naturally.?\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Adjustment Bureau",
"Philip K. Dick",
"Human nature"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4513,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
47345
|
Diuris curta is a species of inedible orchid that flowers from August to early October and is endemic to eastern Australia, its labellum at 10–12 mm (0.39–0.47 in) long has three lobes, with the centre lobe being egg-shaped to wedge-shaped, and both side lobes being oblong to egg-shaped.
|
{
"evidence": [
"Diuris curta_sentence_0",
"Diuris curta_sentence_8",
"Diuris curta_sentence_9",
"Diuris curta_sentence_11"
],
"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:\nDiuris curta is a species of inedible orchid that flowers from August to early October and is endemic to eastern Australia, its labellum at 10–12 mm (0.39–0.47 in) long has three lobes, with the centre lobe being egg-shaped to wedge-shaped, and both side lobes being oblong to egg-shaped.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Diuris curta_sentence_0",
"Diuris curta_sentence_8",
"Diuris curta_sentence_9",
"Diuris curta_sentence_11"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4514,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.845",
"value": "?search="
},
{
"operation": "search",
"time": "18.47",
"value": "Diuris curta"
},
{
"operation": "Now on",
"time": "18.998",
"value": "Diuris curta"
},
{
"operation": "Highlighting",
"time": "24.581",
"value": "Diuris curta_sentence_0"
},
{
"operation": "Highlighting",
"time": "35.054",
"value": "Diuris curta_sentence_8"
},
{
"operation": "Highlighting",
"time": "55.978",
"value": "Diuris curta_sentence_9"
},
{
"operation": "Highlighting deleted",
"time": "103.634",
"value": "Diuris curta_sentence_8"
},
{
"operation": "Highlighting deleted",
"time": "104.41",
"value": "Diuris curta_sentence_9"
},
{
"operation": "hyperlink",
"time": "125.729",
"value": "labellum"
},
{
"operation": "Now on",
"time": "126.514",
"value": "Labellum (botany)"
},
{
"operation": "back-button-clicked",
"time": "132.368",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "132.418",
"value": "Diuris curta"
},
{
"operation": "Highlighting",
"time": "136.349",
"value": "Diuris curta_sentence_8"
},
{
"operation": "Highlighting",
"time": "137.722",
"value": "Diuris curta_sentence_9"
},
{
"operation": "Highlighting",
"time": "170.302",
"value": "Diuris curta_sentence_11"
},
{
"operation": "finish",
"time": "175.364",
"value": "finish"
}
]
}
|
73289
|
Karenne Wood was named as one of the 2015 "Virginia Women in History", an annual program sponsored by the Virginia Foundation for Women.
|
{
"evidence": [
"Virginia Women in History_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:\nKarenne Wood was named as one of the 2015 \"Virginia Women in History\", an annual program sponsored by the Virginia Foundation for Women.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Virginia Women in History_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4515,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.019",
"value": "?search="
},
{
"operation": "search",
"time": "88.62",
"value": "Karenne Wood"
},
{
"operation": "Now on",
"time": "90.113",
"value": "Karenne Wood"
},
{
"operation": "Highlighting",
"time": "92.461",
"value": "Karenne Wood_sentence_2"
},
{
"operation": "Highlighting",
"time": "92.462",
"value": "Karenne Wood_title"
},
{
"operation": "hyperlink",
"time": "99.044",
"value": "Virginia Women in History"
},
{
"operation": "Now on",
"time": "100.888",
"value": "Virginia Women in History"
},
{
"operation": "Highlighting",
"time": "102.324",
"value": "Virginia Women in History_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "126.207",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "126.293",
"value": "Karenne Wood"
},
{
"operation": "Highlighting deleted",
"time": "127.704",
"value": "Karenne Wood_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "127.706",
"value": "Karenne Wood_title"
},
{
"operation": "finish",
"time": "135.038",
"value": "finish"
}
]
}
|
53700
|
In 2012, Algeria national football had joined in eight competitions that resulted in six winnings and two losses, and the competition they had the most winnings was in the African Cup qualification, which is the qualification process for the 2013 Africa Cup of Nations, also known as the 29th edition of the Africa Cup of Nations tournament.
|
{
"evidence": [
"2013 Africa Cup of Nations qualification_sentence_0",
"Algeria national football team_sentence_94",
"Algeria national football team_sentence_95",
"Algeria national football team_sentence_96",
"Algeria national football team_sentence_97"
],
"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 2012, Algeria national football had joined in eight competitions that resulted in six winnings and two losses, and the competition they had the most winnings was in the African Cup qualification, which is the qualification process for the 2013 Africa Cup of Nations, also known as the 29th edition of the Africa Cup of Nations tournament.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"2013 Africa Cup of Nations qualification_sentence_0",
"Algeria national football team_sentence_94",
"Algeria national football team_sentence_95",
"Algeria national football team_sentence_96",
"Algeria national football team_sentence_97"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4516,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.959",
"value": "?search="
},
{
"operation": "search",
"time": "52.22",
"value": "2012 algeria national football"
},
{
"operation": "Now on",
"time": "54.016",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "57.873",
"value": "Algeria national football team results (2010–19)"
},
{
"operation": "Now on",
"time": "58.988",
"value": "Algeria national football team results (2010–19)"
},
{
"operation": "hyperlink",
"time": "62.521",
"value": "3 2012 Algeria national football team results (2010–19)_section_2"
},
{
"operation": "Highlighting",
"time": "71.521",
"value": "Algeria national football team results (2010–19)_sentence_0"
},
{
"operation": "Highlighting",
"time": "71.522",
"value": "Algeria national football team results (2010–19)_title"
},
{
"operation": "hyperlink",
"time": "72.715",
"value": "Algeria national football team"
},
{
"operation": "Now on",
"time": "75.374",
"value": "Algeria national football team"
},
{
"operation": "back-button-clicked",
"time": "104.412",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "104.514",
"value": "Algeria national football team results (2010–19)#2012 Algeria national football team results .282010.E2.80.9319.29 section 2"
},
{
"operation": "Highlighting deleted",
"time": "111.036",
"value": "Algeria national football team results (2010–19)_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "111.04",
"value": "Algeria national football team results (2010–19)_title"
},
{
"operation": "search",
"time": "144.024",
"value": "2012 Algerian Cup Final"
},
{
"operation": "Now on",
"time": "145.138",
"value": "2012 Algerian Cup Final"
},
{
"operation": "search",
"time": "172.234",
"value": "2012 African Cup of Nations qualification"
},
{
"operation": "Now on",
"time": "181.59",
"value": "2012 African Cup of Nations qualification"
},
{
"operation": "search",
"time": "218.113",
"value": "2013 Africa Cup of Nations qualification"
},
{
"operation": "Now on",
"time": "227.937",
"value": "2013 Africa Cup of Nations qualification"
},
{
"operation": "Highlighting",
"time": "247.186",
"value": "2013 Africa Cup of Nations qualification_sentence_0"
},
{
"operation": "hyperlink",
"time": "283.833",
"value": "Algeria"
},
{
"operation": "Now on",
"time": "284.424",
"value": "Algeria national football team"
},
{
"operation": "hyperlink",
"time": "312.144",
"value": "1.5.5 2011–12"
},
{
"operation": "Highlighting",
"time": "423.536",
"value": "Algeria national football team_sentence_94"
},
{
"operation": "Highlighting",
"time": "441.62",
"value": "Algeria national football team_sentence_95"
},
{
"operation": "Highlighting",
"time": "459.465",
"value": "Algeria national football team_sentence_96"
},
{
"operation": "Highlighting",
"time": "467.469",
"value": "Algeria national football team_sentence_97"
},
{
"operation": "finish",
"time": "493.544",
"value": "finish"
}
]
}
|
exfever_2300
|
Jack Twist is an American actor whose older sister is an actress and the daughter of an American film director, who is also a poet.
|
{
"evidence": [
"Jake Gyllenhaal",
"Maggie Gyllenhaal",
"Stephen Gyllenhaal"
],
"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:\nJack Twist is an American actor whose older sister is an actress and the daughter of an American film director, who is also a poet.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Jake Gyllenhaal",
"Maggie Gyllenhaal",
"Stephen Gyllenhaal"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4517,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8685
|
Inhumans also premiere at Universal CityWalk.
|
{
"evidence": [
"Inhumans (TV series)",
"Inhumans"
],
"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:\nInhumans also premiere at Universal CityWalk.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Inhumans (TV series)",
"Inhumans"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4518,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
10784
|
Karla Drenner obtained a Bachelor of Science from West Virginia State College; then in 2011, she completed program for Senior Executives in State and Local Government as a Leadership Fellow.
|
{
"evidence": [
"Karla Drenner_sentence_3",
"Karla Drenner_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:\nKarla Drenner obtained a Bachelor of Science from West Virginia State College; then in 2011, she completed program for Senior Executives in State and Local Government as a Leadership Fellow.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Karla Drenner_sentence_3",
"Karla Drenner_sentence_4"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4519,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.648",
"value": "?search="
},
{
"operation": "search",
"time": "14.055",
"value": "Karla Drenner"
},
{
"operation": "Now on",
"time": "15.792",
"value": "Karla Drenner"
},
{
"operation": "Highlighting",
"time": "28.198",
"value": "Karla Drenner_sentence_3"
},
{
"operation": "Highlighting",
"time": "35.269",
"value": "Karla Drenner_sentence_4"
},
{
"operation": "finish",
"time": "57.293",
"value": "finish"
}
]
}
|
exfever_1494
|
Pablo Picasso was a Spanish theatre designer and painter who was the co-inventor of a technique of art creation, primarily used in art forms such as video, design, and dancing.
|
{
"evidence": [
"Pablo Picasso",
"Collage",
"Visual arts"
],
"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:\nPablo Picasso was a Spanish theatre designer and painter who was the co-inventor of a technique of art creation, primarily used in art forms such as video, design, and dancing.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Pablo Picasso",
"Collage",
"Visual arts"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4520,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8292
|
Francis Sinatra was an American singer and actor who released his first studio album on Columbia Records, catalogue C-112, on March 4, 1946.
|
{
"evidence": [
"Frank Sinatra",
"The Voice of Frank Sinatra"
],
"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:\nFrancis Sinatra was an American singer and actor who released his first studio album on Columbia Records, catalogue C-112, on March 4, 1946.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Frank Sinatra",
"The Voice of Frank Sinatra"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4521,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_2724
|
Céline Marie Claudette Dion is the richest Canadian singer.
|
{
"evidence": [
"Celine Dion",
"Charlemagne, Quebec"
],
"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:\nCéline Marie Claudette Dion is the richest Canadian singer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Celine Dion",
"Charlemagne, Quebec"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4522,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
86450
|
2Spot Communications is a character-design company based in Thailand that has created over 500 characters, and its four most popular ones are Biscuit, Bloody Bunny, P4, and Unsleep Sheep.
|
{
"evidence": [
"2Spot Communications_sentence_11"
],
"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:\n2Spot Communications is a character-design company based in Thailand that has created over 500 characters, and its four most popular ones are Biscuit, Bloody Bunny, P4, and Unsleep Sheep.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"2Spot Communications_sentence_11"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4523,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.027",
"value": "?search="
},
{
"operation": "search",
"time": "11.528",
"value": "2Spot Communications"
},
{
"operation": "Now on",
"time": "12.667",
"value": "2Spot Communications"
},
{
"operation": "Highlighting",
"time": "19.527",
"value": "2Spot Communications_sentence_0"
},
{
"operation": "Highlighting",
"time": "42.22",
"value": "2Spot Communications_sentence_11"
},
{
"operation": "finish",
"time": "70.036",
"value": "finish"
}
]
}
|
exfever_2524
|
Cleopatra VII Philopator was Queen of the Ptolemaic Kingdom of Egypt and was a descendant of an Ancient Macedonian general, historian, and opponent of Alexander the Great.
|
{
"evidence": [
"Cleopatra",
"Ptolemy I Soter"
],
"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:\nCleopatra VII Philopator was Queen of the Ptolemaic Kingdom of Egypt and was a descendant of an Ancient Macedonian general, historian, and opponent of Alexander the Great.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Cleopatra",
"Ptolemy I Soter"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4524,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
5055
|
While at Repton, Fisher met Rosamond Chevallier Forman, who was the daughter of a prime minister, but after graduating in Repton they lived on separate ways.
|
{
"evidence": [
"Geoffrey Fisher_sentence_41"
],
"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:\nWhile at Repton, Fisher met Rosamond Chevallier Forman, who was the daughter of a prime minister, but after graduating in Repton they lived on separate ways.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Geoffrey Fisher_sentence_41"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4525,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "12.424",
"value": "Contains...Rosamond Chevallier Forman"
},
{
"operation": "hyperlink",
"time": "19.374",
"value": "Geoffrey Fisher"
},
{
"operation": "Page search",
"time": "31.974",
"value": "Rosamond Chevallier Forman"
},
{
"operation": "Highlighting",
"time": "39.901",
"value": "Geoffrey Fisher_sentence_41"
},
{
"operation": "Page search",
"time": "58.039",
"value": "graduating"
},
{
"operation": "finish",
"time": "86.323",
"value": "finish"
}
]
}
|
exfever_6623
|
The Shape of Water is an American romantic fantasy film starring an English actress who received a Berlin International Film Festival Silver Bear for Best Dancer.
|
{
"evidence": [
"The Shape of Water (film)",
"Sally Hawkins"
],
"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 Shape of Water is an American romantic fantasy film starring an English actress who received a Berlin International Film Festival Silver Bear for Best Dancer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Shape of Water (film)",
"Sally Hawkins"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4526,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_7967
|
Laura Elizabeth Dern rose to prominence in David Lynch's 1986 neo-noir mystery thriller film, starring an English actress and filmmaker.
|
{
"evidence": [
"Laura Dern",
"Blue Velvet (film)",
"Laura Dern"
],
"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:\nLaura Elizabeth Dern rose to prominence in David Lynch's 1986 neo-noir mystery thriller film, starring an English actress and filmmaker.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Laura Dern",
"Blue Velvet (film)",
"Laura Dern"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4527,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
34013
|
Earl Steven Karr, responsible for the May-June 1984 Midwest pipe bombings that took place in Ohio, was arrested in Iowa while after a pipe bomb detonated in the trunk of his car.
|
{
"evidence": [
"1984 Midwest pipe bombings_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:\nEarl Steven Karr, responsible for the May-June 1984 Midwest pipe bombings that took place in Ohio, was arrested in Iowa while after a pipe bomb detonated in the trunk of his car.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"1984 Midwest pipe bombings_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4528,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.953",
"value": "?search="
},
{
"operation": "search",
"time": "13.92",
"value": "Earl Steven Karr"
},
{
"operation": "Now on",
"time": "15.39",
"value": "1984 Midwest pipe bombings"
},
{
"operation": "Highlighting",
"time": "31.363",
"value": "1984 Midwest pipe bombings_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "32.997",
"value": "1984 Midwest pipe bombings_sentence_0"
},
{
"operation": "Highlighting",
"time": "34.408",
"value": "1984 Midwest pipe bombings_sentence_1"
},
{
"operation": "finish",
"time": "52.285",
"value": "finish"
}
]
}
|
exfever_6636
|
Octavia Lenora Spencer got her breakthrough in 2011 when she played a maid in a 2011 period comedy-drama film written and directed by Tate Taylor.
|
{
"evidence": [
"Octavia Spencer",
"The Help (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:\nOctavia Lenora Spencer got her breakthrough in 2011 when she played a maid in a 2011 period comedy-drama film written and directed by Tate Taylor.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Octavia Spencer",
"The Help (film)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4529,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
44401
|
The Langham Place (All Souls) serves a population of 8,028; while St John's Wood (St John) serves a population of 8,533.
|
{
"evidence": [
"All Souls Church, Langham Place_sentence_3",
"St John's Wood Church_sentence_11"
],
"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 Langham Place (All Souls) serves a population of 8,028; while St John's Wood (St John) serves a population of 8,533.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"All Souls Church, Langham Place_sentence_3",
"St John's Wood Church_sentence_11"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4530,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.772",
"value": "?search="
},
{
"operation": "search",
"time": "9.003",
"value": "Contains...The Langham Place"
},
{
"operation": "Now on",
"time": "10.582",
"value": "undefined"
},
{
"operation": "search",
"time": "27.005",
"value": "All Souls"
},
{
"operation": "Now on",
"time": "27.979",
"value": "All Souls"
},
{
"operation": "search",
"time": "37.545",
"value": "St John's Wood"
},
{
"operation": "Now on",
"time": "40.822",
"value": "St John's Wood"
},
{
"operation": "hyperlink",
"time": "57.46",
"value": "St John's Wood, Queensland"
},
{
"operation": "Now on",
"time": "59.549",
"value": "St Johns Wood, Queensland"
},
{
"operation": "back-button-clicked",
"time": "65.71",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "65.849",
"value": "St John's Wood"
},
{
"operation": "Page search",
"time": "70.593",
"value": "population"
},
{
"operation": "search",
"time": "85.621",
"value": "Langham Place"
},
{
"operation": "Now on",
"time": "86.316",
"value": "Langham Place"
},
{
"operation": "hyperlink",
"time": "91.706",
"value": "Langham Place, London"
},
{
"operation": "Now on",
"time": "92.679",
"value": "Langham Place, London"
},
{
"operation": "hyperlink",
"time": "98.404",
"value": "1.3 All Souls Church Langham Place, London_section_3"
},
{
"operation": "hyperlink",
"time": "101.259",
"value": "All Souls Church, Langham Place"
},
{
"operation": "Now on",
"time": "102.149",
"value": "All Souls Church, Langham Place"
},
{
"operation": "Page search",
"time": "122.801",
"value": "serves"
},
{
"operation": "Page search",
"time": "136.254",
"value": "population"
},
{
"operation": "hyperlink",
"time": "143.114",
"value": "Langham Place"
},
{
"operation": "Now on",
"time": "143.621",
"value": "Langham Place, London"
},
{
"operation": "hyperlink",
"time": "150.921",
"value": "All Souls Church"
},
{
"operation": "Now on",
"time": "151.451",
"value": "All Souls Church, Langham Place"
},
{
"operation": "Highlighting",
"time": "155.21",
"value": "All Souls Church, Langham Place_sentence_3"
},
{
"operation": "search",
"time": "195.798",
"value": "St. John's Wood Church Grounds"
},
{
"operation": "Now on",
"time": "196.82",
"value": "St. John's Wood Church Grounds"
},
{
"operation": "back-button-clicked",
"time": "201.559",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "201.716",
"value": "All Souls Church, Langham Place"
},
{
"operation": "search",
"time": "209.718",
"value": "St John's Wood Church"
},
{
"operation": "Now on",
"time": "210.936",
"value": "St John's Wood Church"
},
{
"operation": "Highlighting",
"time": "226.693",
"value": "St John's Wood Church_sentence_11"
},
{
"operation": "hyperlink",
"time": "233.666",
"value": "List of churches in London"
},
{
"operation": "Now on",
"time": "243.724",
"value": "List of churches in London"
},
{
"operation": "Page search",
"time": "273.713",
"value": "langham place"
},
{
"operation": "Page search",
"time": "278.117",
"value": "all souls"
},
{
"operation": "Highlighting",
"time": "318.124",
"value": "List of churches in London_cell_3_19_0"
},
{
"operation": "Highlighting",
"time": "318.127",
"value": "List of churches in London_title"
},
{
"operation": "Highlighting",
"time": "318.129",
"value": "List of churches in London_section_3"
},
{
"operation": "Highlighting",
"time": "318.13",
"value": "List of churches in London_section_9"
},
{
"operation": "Highlighting",
"time": "318.139",
"value": "List of churches in London_header_cell_3_0_0"
},
{
"operation": "Highlighting",
"time": "319.807",
"value": "List of churches in London_cell_3_19_2"
},
{
"operation": "Highlighting",
"time": "319.818",
"value": "List of churches in London_header_cell_3_0_2"
},
{
"operation": "hyperlink",
"time": "323.281",
"value": "All Souls"
},
{
"operation": "Now on",
"time": "324.662",
"value": "All Souls' Day"
},
{
"operation": "back-button-clicked",
"time": "327.552",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "329.951",
"value": "List of churches in London"
},
{
"operation": "back-button-clicked",
"time": "353.02",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "353.027",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "353.232",
"value": "St John's Wood Church"
},
{
"operation": "search",
"time": "364.905",
"value": "St John's Wood"
},
{
"operation": "Now on",
"time": "366.187",
"value": "St John's Wood"
},
{
"operation": "Page search",
"time": "379.237",
"value": "8,533"
},
{
"operation": "search",
"time": "412.418",
"value": "St John's Wood Church"
},
{
"operation": "Now on",
"time": "413.392",
"value": "St John's Wood Church"
},
{
"operation": "Highlighting",
"time": "421.398",
"value": "St John's Wood Church_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "423.102",
"value": "St John's Wood Church_sentence_0"
},
{
"operation": "finish",
"time": "426.458",
"value": "finish"
}
]
}
|
90179
|
Sepia faurei, a species of cuttlefish, is only found in the Pacific ocean and lives to depths of 168 meters.
|
{
"evidence": [
"Sepia faurei_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:\nSepia faurei, a species of cuttlefish, is only found in the Pacific ocean and lives to depths of 168 meters.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Sepia faurei_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4531,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.016",
"value": "?search="
},
{
"operation": "search",
"time": "14.048",
"value": "Sepia faurei"
},
{
"operation": "Now on",
"time": "14.803",
"value": "Sepia faurei"
},
{
"operation": "Highlighting",
"time": "21.399",
"value": "Sepia faurei_sentence_0"
},
{
"operation": "finish",
"time": "35.934",
"value": "finish"
}
]
}
|
exfever_6528
|
Julian Fellowes is an English director who has a net worth of 10 million dollars.
|
{
"evidence": [
"Gareth Neame",
"Downton Abbey"
],
"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:\nJulian Fellowes is an English director who has a net worth of 10 million dollars.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Gareth Neame",
"Downton Abbey"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4532,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
61460
|
Schnaus is a former municipality in the district of Imboden. in the Graubünden canton with an erratic number of population; 126(1850), 123(1900), 134(1950), 79(1990) and 99(2000).
|
{
"evidence": [
"Schnaus_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:\nSchnaus is a former municipality in the district of Imboden. in the Graubünden canton with an erratic number of population; 126(1850), 123(1900), 134(1950), 79(1990) and 99(2000).\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Schnaus_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4533,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.421",
"value": "?search="
},
{
"operation": "search",
"time": "19.252",
"value": "Schnaus"
},
{
"operation": "Now on",
"time": "20.009",
"value": "Schnaus"
},
{
"operation": "Highlighting",
"time": "21.696",
"value": "Schnaus_sentence_0"
},
{
"operation": "finish",
"time": "81.401",
"value": "finish"
}
]
}
|
46773
|
Nguyễn Khuyến was smart but did not do good on tests but he still passed and got a Masters on the Cử nhân degree.
|
{
"evidence": [
"Nguyễn Khuyến_sentence_7",
"Nguyễn Khuyến_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:\nNguyễn Khuyến was smart but did not do good on tests but he still passed and got a Masters on the Cử nhân degree.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Nguyễn Khuyến_sentence_7",
"Nguyễn Khuyến_sentence_8"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4534,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.781",
"value": "?search="
},
{
"operation": "search",
"time": "15.302",
"value": "Nguyễn Khuyến"
},
{
"operation": "Now on",
"time": "16.103",
"value": "Nguyễn Khuyến"
},
{
"operation": "Highlighting",
"time": "23.863",
"value": "Nguyễn Khuyến_sentence_0"
},
{
"operation": "Highlighting",
"time": "194.395",
"value": "Nguyễn Khuyến_sentence_7"
},
{
"operation": "Highlighting",
"time": "205.126",
"value": "Nguyễn Khuyến_sentence_8"
},
{
"operation": "Highlighting deleted",
"time": "224.336",
"value": "Nguyễn Khuyến_sentence_0"
},
{
"operation": "finish",
"time": "229.961",
"value": "finish"
}
]
}
|
exfever_4187
|
Splash is the third American fantasy romantic comedy film directed by Ron Howard.
|
{
"evidence": [
"Splash (film)",
"Tom Hanks"
],
"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:\nSplash is the third American fantasy romantic comedy film directed by Ron Howard.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Splash (film)",
"Tom Hanks"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4535,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
25455
|
Cai Jinbiao, a former Chinese international footballer who played neither as a defender nor a striker, won the World Cup Qualifier in 1980-1982 with twelve appearances and one goal.
|
{
"evidence": [
"Cai Jinbiao_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:\nCai Jinbiao, a former Chinese international footballer who played neither as a defender nor a striker, won the World Cup Qualifier in 1980-1982 with twelve appearances and one goal.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Cai Jinbiao_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4536,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.582",
"value": "?search="
},
{
"operation": "search",
"time": "10.995",
"value": "Cai Jinbiao"
},
{
"operation": "Now on",
"time": "11.936",
"value": "Cai Jinbiao"
},
{
"operation": "Highlighting",
"time": "17.067",
"value": "Cai Jinbiao_sentence_0"
},
{
"operation": "finish",
"time": "24.822",
"value": "finish"
}
]
}
|
exfever_2968
|
The Yellow Flicker Beat is a song that is electro-pop and a loosely defined style of pop music influenced by pop art's integration of high and low culture.
|
{
"evidence": [
"Yellow Flicker Beat",
"Art pop"
],
"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 Yellow Flicker Beat is a song that is electro-pop and a loosely defined style of pop music influenced by pop art's integration of high and low culture.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Yellow Flicker Beat",
"Art pop"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4537,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3475
|
Goat Island is a small island in the middle of a group of three waterfalls spanning the border between the province of Ontario in Canada and the state of New York in the United States.
|
{
"evidence": [
"Goat Island (New York)",
"Niagara Falls"
],
"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:\nGoat Island is a small island in the middle of a group of three waterfalls spanning the border between the province of Ontario in Canada and the state of New York in the United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Goat Island (New York)",
"Niagara Falls"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4538,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5051
|
William Richard Tolbert Jr. was the 20th President of Liberia who was killed in a coup d'état led by a Liberian politician who served as the Liberian leader from 1980 to 1997.
|
{
"evidence": [
"William R. Tolbert Jr.",
"Samuel Doe"
],
"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:\nWilliam Richard Tolbert Jr. was the 20th President of Liberia who was killed in a coup d'état led by a Liberian politician who served as the Liberian leader from 1980 to 1997.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"William R. Tolbert Jr.",
"Samuel Doe"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4539,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
76181
|
The wingspan of Ditrigona lineata is smaller for females than males, their forewings and hindwings are lustrous white and is found in China whose capital is Beijing.
|
{
"evidence": [
"Ditrigona lineata_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:\nThe wingspan of Ditrigona lineata is smaller for females than males, their forewings and hindwings are lustrous white and is found in China whose capital is Beijing.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Ditrigona lineata_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4540,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.756",
"value": "?search="
},
{
"operation": "search",
"time": "10.521",
"value": "Ditrigona lineata"
},
{
"operation": "Now on",
"time": "11.379",
"value": "Ditrigona lineata"
},
{
"operation": "Highlighting",
"time": "17.541",
"value": "Ditrigona lineata_sentence_3"
},
{
"operation": "finish",
"time": "30.892",
"value": "finish"
}
]
}
|
92862
|
Ceranemota semifasciata have gray forewings, a white base with a black do in the middle, and towards the inner margin a small amount of black and white scales.
|
{
"evidence": [
"Ceranemota semifasciata_sentence_4",
"Ceranemota semifasciata_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:\nCeranemota semifasciata have gray forewings, a white base with a black do in the middle, and towards the inner margin a small amount of black and white scales.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Ceranemota semifasciata_sentence_4",
"Ceranemota semifasciata_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4541,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.689",
"value": "?search="
},
{
"operation": "search",
"time": "9.589",
"value": "Ceranemota semifasciata"
},
{
"operation": "Now on",
"time": "11.069",
"value": "Ceranemota semifasciata"
},
{
"operation": "Highlighting",
"time": "13.701",
"value": "Ceranemota semifasciata_sentence_4"
},
{
"operation": "Highlighting",
"time": "16.594",
"value": "Ceranemota semifasciata_sentence_5"
},
{
"operation": "finish",
"time": "37.682",
"value": "finish"
}
]
}
|
exfever_8080
|
To Kill a Mockingbird is a novel by the American author Harper Lee, loosely based on events that occurred in the county seat of Monroe County.
|
{
"evidence": [
"To Kill a Mockingbird",
"Monroeville, Alabama"
],
"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:\nTo Kill a Mockingbird is a novel by the American author Harper Lee, loosely based on events that occurred in the county seat of Monroe County.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"To Kill a Mockingbird",
"Monroeville, Alabama"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4542,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
89855
|
Matt Watkins signed a five-year contract with the Phoenix Coyotes, but was assigned to the Portland Pirates.
|
{
"evidence": [
"Matt Watkins_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:\nMatt Watkins signed a five-year contract with the Phoenix Coyotes, but was assigned to the Portland Pirates.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Matt Watkins_sentence_6"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4543,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.289",
"value": "?search="
},
{
"operation": "search",
"time": "9.488",
"value": "Matt Watkins"
},
{
"operation": "Now on",
"time": "10.408",
"value": "Matt Watkins"
},
{
"operation": "Highlighting",
"time": "23.984",
"value": "Matt Watkins_sentence_6"
},
{
"operation": "Highlighting",
"time": "34.284",
"value": "Matt Watkins_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "47.245",
"value": "Matt Watkins_sentence_7"
},
{
"operation": "Highlighting",
"time": "77.142",
"value": "Matt Watkins_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "82.538",
"value": "Matt Watkins_sentence_7"
},
{
"operation": "Highlighting",
"time": "172.786",
"value": "Matt Watkins_sentence_7"
},
{
"operation": "Highlighting",
"time": "173.34",
"value": "Matt Watkins_sentence_8"
},
{
"operation": "Highlighting",
"time": "173.869",
"value": "Matt Watkins_sentence_9"
},
{
"operation": "Highlighting",
"time": "174.385",
"value": "Matt Watkins_sentence_10"
},
{
"operation": "Highlighting",
"time": "174.919",
"value": "Matt Watkins_sentence_11"
},
{
"operation": "Highlighting",
"time": "175.895",
"value": "Matt Watkins_sentence_12"
},
{
"operation": "Highlighting",
"time": "176.432",
"value": "Matt Watkins_sentence_13"
},
{
"operation": "Highlighting",
"time": "177.575",
"value": "Matt Watkins_sentence_14"
},
{
"operation": "Highlighting",
"time": "178.773",
"value": "Matt Watkins_cell_0_8_1"
},
{
"operation": "Highlighting",
"time": "178.78",
"value": "Matt Watkins_header_cell_0_8_0"
},
{
"operation": "Highlighting",
"time": "178.78",
"value": "Matt Watkins_header_cell_0_0_0"
},
{
"operation": "Highlighting deleted",
"time": "188.621",
"value": "Matt Watkins_sentence_14"
},
{
"operation": "Highlighting deleted",
"time": "189.038",
"value": "Matt Watkins_sentence_13"
},
{
"operation": "Highlighting deleted",
"time": "189.493",
"value": "Matt Watkins_sentence_11"
},
{
"operation": "Highlighting deleted",
"time": "190.976",
"value": "Matt Watkins_sentence_12"
},
{
"operation": "Highlighting deleted",
"time": "191.542",
"value": "Matt Watkins_sentence_9"
},
{
"operation": "Highlighting deleted",
"time": "192.076",
"value": "Matt Watkins_sentence_10"
},
{
"operation": "Highlighting deleted",
"time": "192.75",
"value": "Matt Watkins_sentence_8"
},
{
"operation": "Highlighting deleted",
"time": "193.242",
"value": "Matt Watkins_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "194.12",
"value": "Matt Watkins_cell_0_8_1"
},
{
"operation": "Highlighting deleted",
"time": "194.129",
"value": "Matt Watkins_header_cell_0_8_0"
},
{
"operation": "Highlighting deleted",
"time": "194.131",
"value": "Matt Watkins_header_cell_0_0_0"
},
{
"operation": "finish",
"time": "210.316",
"value": "finish"
}
]
}
|
exfever_2002
|
To Die For is a 1995 film based on a novel of the same name by an American novelist and journalist born on November 5, 1953.
|
{
"evidence": [
"To Die For",
"Joyce Maynard"
],
"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:\nTo Die For is a 1995 film based on a novel of the same name by an American novelist and journalist born on November 5, 1953.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"To Die For",
"Joyce Maynard"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4544,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
41213
|
The Order of the Netherlands Lion is a Dutch order of chivalry founded by Major General H. Morsink and was until recently awarded upon eminent individuals from all walks of life.
|
{
"evidence": [
"Order of the Netherlands Lion_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 Order of the Netherlands Lion is a Dutch order of chivalry founded by Major General H. Morsink and was until recently awarded upon eminent individuals from all walks of life.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Order of the Netherlands Lion_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4545,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.637",
"value": "?search="
},
{
"operation": "search",
"time": "8.946",
"value": "The Order of the Netherlands Lion"
},
{
"operation": "Now on",
"time": "11.079",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "45.188",
"value": "Order of Orange-Nassau"
},
{
"operation": "Now on",
"time": "45.963",
"value": "Order of Orange-Nassau"
},
{
"operation": "hyperlink",
"time": "65.662",
"value": "Order of the Netherlands Lion"
},
{
"operation": "Now on",
"time": "66.091",
"value": "Order of the Netherlands Lion"
},
{
"operation": "Highlighting",
"time": "75.064",
"value": "Order of the Netherlands Lion_sentence_2"
},
{
"operation": "Page search",
"time": "93.19",
"value": "Morsink"
},
{
"operation": "finish",
"time": "131.972",
"value": "finish"
}
]
}
|
exfever_1612
|
Tobey Maguire is an American actor and film producer who stars in an American teen fantasy comedy-drama film written, co-produced, and directed by Gary Ross.
|
{
"evidence": [
"Tobey Maguire",
"Pleasantville (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:\nTobey Maguire is an American actor and film producer who stars in an American teen fantasy comedy-drama film written, co-produced, and directed by Gary Ross.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tobey Maguire",
"Pleasantville (film)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4546,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5544
|
Bharat Arun was a part of a domestic cricket team thatplays in a domestic Twenty20 cricket championship in India.
|
{
"evidence": [
"Bharat Arun",
"Tamil Nadu cricket team",
"Syed Mushtaq Ali Trophy"
],
"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:\nBharat Arun was a part of a domestic cricket team thatplays in a domestic Twenty20 cricket championship in India.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Bharat Arun",
"Tamil Nadu cricket team",
"Syed Mushtaq Ali Trophy"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4547,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
77873
|
In the 1997 Conference, an Alabama school club was with the D-II Independent, while in the 1998 Conference, it was changed to MIFC.
|
{
"evidence": [
"NCAA Division II independent schools_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 1997 Conference, an Alabama school club was with the D-II Independent, while in the 1998 Conference, it was changed to MIFC.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"NCAA Division II independent schools_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4548,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.873",
"value": "?search="
},
{
"operation": "search",
"time": "11.914",
"value": "1997 Conference alabama"
},
{
"operation": "Now on",
"time": "14.113",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "20.661",
"value": "Southwestern Athletic Conference"
},
{
"operation": "Now on",
"time": "24.63",
"value": "Southwestern Athletic Conference"
},
{
"operation": "search",
"time": "52.011",
"value": "Alabama school club was with the D-II Independent"
},
{
"operation": "Now on",
"time": "55.104",
"value": "undefined"
},
{
"operation": "search",
"time": "64.566",
"value": "1997 conference D-II"
},
{
"operation": "Now on",
"time": "65.716",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "72.311",
"value": "NCAA Division II College basketball_section_10"
},
{
"operation": "Now on",
"time": "74.052",
"value": "College basketball#NCAA Division II College basketball section 10"
},
{
"operation": "search",
"time": "110.373",
"value": "1998 Conference USA Men's Basketball Tournament"
},
{
"operation": "Now on",
"time": "111.632",
"value": "1998 Conference USA Men's Basketball Tournament"
},
{
"operation": "Page search",
"time": "117.632",
"value": "MIFC"
},
{
"operation": "hyperlink",
"time": "127.792",
"value": "1998 NCAA Tournament"
},
{
"operation": "Now on",
"time": "130.302",
"value": "1998 NCAA Division I Men's Basketball Tournament"
},
{
"operation": "hyperlink",
"time": "141.596",
"value": "1998 NCAA Division II Men's Basketball Tournament"
},
{
"operation": "Now on",
"time": "142.602",
"value": "1998 NCAA Division II Men's Basketball Tournament"
},
{
"operation": "hyperlink",
"time": "153.349",
"value": "Division II"
},
{
"operation": "Now on",
"time": "154.817",
"value": "NCAA Division II"
},
{
"operation": "search",
"time": "161.328",
"value": "D-II Independent"
},
{
"operation": "Now on",
"time": "163.102",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "166.877",
"value": "Independents NCAA Division III_section_4"
},
{
"operation": "Now on",
"time": "167.858",
"value": "NCAA Division III#Independents NCAA Division III section 4"
},
{
"operation": "hyperlink",
"time": "177.402",
"value": "Division II"
},
{
"operation": "Now on",
"time": "179.08",
"value": "NCAA Division II"
},
{
"operation": "Page search",
"time": "181.206",
"value": "inde"
},
{
"operation": "hyperlink",
"time": "186.486",
"value": "3.2 Independents NCAA Division II_section_9"
},
{
"operation": "hyperlink",
"time": "189.522",
"value": "Division II Independent schools"
},
{
"operation": "Now on",
"time": "192.004",
"value": "NCAA Division II independent schools"
},
{
"operation": "Highlighting",
"time": "255.699",
"value": "NCAA Division II independent schools_sentence_0"
},
{
"operation": "finish",
"time": "259.318",
"value": "finish"
}
]
}
|
exfever_7018
|
Streptococcus belongs to the family of Gram-positive bacteria, placed within the order Salmonella.
|
{
"evidence": [
"Streptococcus",
"Streptococcaceae"
],
"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:\nStreptococcus belongs to the family of Gram-positive bacteria, placed within the order Salmonella.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Streptococcus",
"Streptococcaceae"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4549,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1884
|
Robert Singer's fictionalized version appears in the season 6 episode The French Mistake, played by an American actor who appeared in the animated series created by a marine science educator.
|
{
"evidence": [
"Robert Singer (producer)",
"Brian Doyle-Murray",
"SpongeBob SquarePants"
],
"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:\nRobert Singer's fictionalized version appears in the season 6 episode The French Mistake, played by an American actor who appeared in the animated series created by a marine science educator.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Robert Singer (producer)",
"Brian Doyle-Murray",
"SpongeBob SquarePants"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4550,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
84048
|
PTC's network included the Philadelphia trolleybus system, the PTC replaced its aging fleet of vehicles but its workers engaged in a wildcat strike which are not typically criminal offenses.
|
{
"evidence": [
"Philadelphia Transportation Company_sentence_6",
"Philadelphia Transportation Company_sentence_7",
"Philadelphia Transportation Company_sentence_9",
"Wildcat strike action_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:\nPTC's network included the Philadelphia trolleybus system, the PTC replaced its aging fleet of vehicles but its workers engaged in a wildcat strike which are not typically criminal offenses.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Philadelphia Transportation Company_sentence_6",
"Philadelphia Transportation Company_sentence_7",
"Philadelphia Transportation Company_sentence_9",
"Wildcat strike action_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4551,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.11",
"value": "?search="
},
{
"operation": "search",
"time": "16.67",
"value": "PTC"
},
{
"operation": "Now on",
"time": "18.18",
"value": "PTC"
},
{
"operation": "hyperlink",
"time": "35.849",
"value": "Pennsylvania Turnpike Commission"
},
{
"operation": "hyperlink",
"time": "39.181",
"value": "Pennsylvania Turnpike Commission"
},
{
"operation": "Now on",
"time": "43.415",
"value": "Pennsylvania Turnpike Commission"
},
{
"operation": "search",
"time": "77.541",
"value": "Philadelphia Transportation Company"
},
{
"operation": "Now on",
"time": "78.418",
"value": "Philadelphia Transportation Company"
},
{
"operation": "Highlighting",
"time": "83.654",
"value": "Philadelphia Transportation Company_sentence_6"
},
{
"operation": "Highlighting",
"time": "95.012",
"value": "Philadelphia Transportation Company_sentence_7"
},
{
"operation": "Highlighting",
"time": "106.606",
"value": "Philadelphia Transportation Company_sentence_9"
},
{
"operation": "hyperlink",
"time": "107.832",
"value": "wildcat strike"
},
{
"operation": "Now on",
"time": "110.712",
"value": "Wildcat strike action"
},
{
"operation": "Highlighting",
"time": "114.244",
"value": "Wildcat strike action_sentence_1"
},
{
"operation": "finish",
"time": "130.726",
"value": "finish"
}
]
}
|
exfever_8861
|
Pulp Fiction is a 1995 comedy crime film starring an actor who also starred in the Die Hard franchise as a fictional character and the main protagonist of the film series, based on Roderick Thorp's action novel, Nothing Lasts Forever.
|
{
"evidence": [
"Pulp Fiction",
"Bruce Willis",
"John McClane"
],
"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:\nPulp Fiction is a 1995 comedy crime film starring an actor who also starred in the Die Hard franchise as a fictional character and the main protagonist of the film series, based on Roderick Thorp's action novel, Nothing Lasts Forever.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Pulp Fiction",
"Bruce Willis",
"John McClane"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4552,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8128
|
Ronald Belford "Bon" Scott was a Scottish-born French singer and songwriter who lived in the capital and most-populous city of the Australian state of Victoria for four years before settling in Fremantle, Western Australia.
|
{
"evidence": [
"Bon Scott",
"Melbourne"
],
"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:\nRonald Belford \"Bon\" Scott was a Scottish-born French singer and songwriter who lived in the capital and most-populous city of the Australian state of Victoria for four years before settling in Fremantle, Western Australia.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Bon Scott",
"Melbourne"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4553,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
90927
|
Sunny Odogwu traveled abroad for further studies between 1954 and 1958 and took up a position of assistant director with C.T. Bowring upon his return to Nigeria.
|
{
"evidence": [
"Sunny Odogwu_sentence_5",
"Sunny Odogwu_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:\nSunny Odogwu traveled abroad for further studies between 1954 and 1958 and took up a position of assistant director with C.T. Bowring upon his return to Nigeria.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Sunny Odogwu_sentence_5",
"Sunny Odogwu_sentence_6"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4554,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "11.869",
"value": "?search="
},
{
"operation": "search",
"time": "17.563",
"value": "Sunny Odogwu"
},
{
"operation": "Now on",
"time": "19.194",
"value": "Sunny Odogwu"
},
{
"operation": "Highlighting",
"time": "47.366",
"value": "Sunny Odogwu_sentence_5"
},
{
"operation": "Highlighting",
"time": "52.176",
"value": "Sunny Odogwu_sentence_6"
},
{
"operation": "finish",
"time": "59.062",
"value": "finish"
}
]
}
|
exfever_3342
|
Winona Ryder gained attention with her performance in Tim Burton's 1988 fantasy comedy film starring an Australian actor.
|
{
"evidence": [
"Winona Ryder",
"Beetlejuice",
"Michael Keaton"
],
"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:\nWinona Ryder gained attention with her performance in Tim Burton's 1988 fantasy comedy film starring an Australian actor.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Winona Ryder",
"Beetlejuice",
"Michael Keaton"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4555,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3870
|
The Vampire Diaries is an American fantasy teen drama television series that was developed by an American screenwriter, director, and producer best known for developing and writing the slasher film Scream.
|
{
"evidence": [
"The Vampire Diaries",
"Kevin Williamson (screenwriter)"
],
"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 Vampire Diaries is an American fantasy teen drama television series that was developed by an American screenwriter, director, and producer best known for developing and writing the slasher film Scream.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Vampire Diaries",
"Kevin Williamson (screenwriter)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4556,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3409
|
Mike Hamlin and the American comedian, television personality, and actress who acted in the 1996 romantic black comedy film, launched the record label Eleveneleven in 2010.
|
{
"evidence": [
"Eleveneleven",
"Ellen DeGeneres",
"Mr. Wrong"
],
"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:\nMike Hamlin and the American comedian, television personality, and actress who acted in the 1996 romantic black comedy film, launched the record label Eleveneleven in 2010.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Eleveneleven",
"Ellen DeGeneres",
"Mr. Wrong"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4557,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
36125
|
1966 German Grand Prix had 30 drivers, one of whom is Jim Clark (a American Formula One racing driver ).
|
{
"evidence": [
"Jim Clark_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:\n1966 German Grand Prix had 30 drivers, one of whom is Jim Clark (a American Formula One racing driver ).\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Jim Clark_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4558,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.246",
"value": "?search="
},
{
"operation": "search",
"time": "10.978",
"value": "1966 German Grand Prix"
},
{
"operation": "Now on",
"time": "12.476",
"value": "1966 German Grand Prix"
},
{
"operation": "Highlighting",
"time": "41.755",
"value": "1966 German Grand Prix_cell_0_1_2"
},
{
"operation": "Highlighting",
"time": "41.757",
"value": "1966 German Grand Prix_title"
},
{
"operation": "Highlighting",
"time": "41.758",
"value": "1966 German Grand Prix_section_1"
},
{
"operation": "Highlighting",
"time": "41.758",
"value": "1966 German Grand Prix_section_2"
},
{
"operation": "Highlighting",
"time": "41.829",
"value": "1966 German Grand Prix_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "41.83",
"value": "1966 German Grand Prix_header_cell_0_0_2"
},
{
"operation": "Highlighting deleted",
"time": "43.656",
"value": "1966 German Grand Prix_cell_0_1_2"
},
{
"operation": "Highlighting deleted",
"time": "43.659",
"value": "1966 German Grand Prix_title"
},
{
"operation": "Highlighting deleted",
"time": "43.665",
"value": "1966 German Grand Prix_section_1"
},
{
"operation": "Highlighting deleted",
"time": "43.667",
"value": "1966 German Grand Prix_section_2"
},
{
"operation": "Highlighting deleted",
"time": "43.734",
"value": "1966 German Grand Prix_header_cell_0_1_0"
},
{
"operation": "Highlighting deleted",
"time": "43.736",
"value": "1966 German Grand Prix_header_cell_0_0_2"
},
{
"operation": "hyperlink",
"time": "44.863",
"value": "Jim Clark"
},
{
"operation": "Now on",
"time": "46.411",
"value": "Jim Clark"
},
{
"operation": "Highlighting",
"time": "49.512",
"value": "Jim Clark_sentence_1"
},
{
"operation": "finish",
"time": "61.447",
"value": "finish"
}
]
}
|
exfever_1807
|
Elisabeth Sladen was an English actress best known in the television series Doctor Who as the investigative journalist dedicated to protecting Earth from alien invasions.
|
{
"evidence": [
"Elisabeth Sladen",
"Sarah Jane Smith"
],
"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:\nElisabeth Sladen was an English actress best known in the television series Doctor Who as the investigative journalist dedicated to protecting Earth from alien invasions.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Elisabeth Sladen",
"Sarah Jane Smith"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4559,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_7945
|
Aubrey Drake Graham is a Canadian rapper, singer, songwriter, and actor who released his debut studio album on June 15, 2011, by an American record label founded by rapper Lil Wayne.
|
{
"evidence": [
"Drake (musician)",
"Thank Me Later",
"Young Money Entertainment"
],
"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:\nAubrey Drake Graham is a Canadian rapper, singer, songwriter, and actor who released his debut studio album on June 15, 2011, by an American record label founded by rapper Lil Wayne.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Drake (musician)",
"Thank Me Later",
"Young Money Entertainment"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4560,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8456
|
The Daily Show is the first American late-night talk and news television program that airs on Comedy Central.
|
{
"evidence": [
"The Daily Show",
"News satire",
"Parody"
],
"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 Daily Show is the first American late-night talk and news television program that airs on Comedy Central.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Daily Show",
"News satire",
"Parody"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4561,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
25259
|
Dead Doll featured Natalie Davis using a taser on Sara who tries to escape and successfully gets away from Natalie who wants to drugs her.
|
{
"evidence": [
"Dead Doll_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:\nDead Doll featured Natalie Davis using a taser on Sara who tries to escape and successfully gets away from Natalie who wants to drugs her.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Dead Doll_sentence_5"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4562,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.587",
"value": "?search="
},
{
"operation": "search",
"time": "10.209",
"value": "Dead Doll"
},
{
"operation": "Now on",
"time": "11.133",
"value": "Dead Doll"
},
{
"operation": "Highlighting",
"time": "183.768",
"value": "Dead Doll_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "186.511",
"value": "Dead Doll_sentence_0"
},
{
"operation": "Page search",
"time": "190.62",
"value": "natalie"
},
{
"operation": "Highlighting",
"time": "193.226",
"value": "Dead Doll_sentence_3"
},
{
"operation": "Highlighting",
"time": "198.721",
"value": "Dead Doll_sentence_4"
},
{
"operation": "Highlighting",
"time": "207.599",
"value": "Dead Doll_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "208.843",
"value": "Dead Doll_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "210.87",
"value": "Dead Doll_sentence_3"
},
{
"operation": "finish",
"time": "217.719",
"value": "finish"
}
]
}
|
exfever_1572
|
Ronald Wilson Reagan was a member of one of the two major contemporary parties who served as the 40th president of the United States.
|
{
"evidence": [
"Ronald Reagan",
"Republican Party (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:\nRonald Wilson Reagan was a member of one of the two major contemporary parties who served as the 40th president of the United States.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Ronald Reagan",
"Republican Party (United States)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4563,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_7041
|
Unapologetic is the tenth studio album by Rihanna who collaborated with a British record producer who worked with an Australian singer and songwriter who is best known for a cover version of "You Don't Own Me".
|
{
"evidence": [
"Unapologetic",
"Parker Ighile",
"Grace (Australian singer)"
],
"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:\nUnapologetic is the tenth studio album by Rihanna who collaborated with a British record producer who worked with an Australian singer and songwriter who is best known for a cover version of \"You Don't Own Me\".\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Unapologetic",
"Parker Ighile",
"Grace (Australian singer)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4564,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_4084
|
Treasure is a song by American singer and songwriter Bruno Mars produced by an American songwriting and record production team consisting of Bruno Mars, Philip Lawrence and Ari Levine established in a state in the Western United States which is the most populous and the third-largest United State.
|
{
"evidence": [
"Treasure (Bruno Mars song)",
"The Smeezingtons",
"California"
],
"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:\nTreasure is a song by American singer and songwriter Bruno Mars produced by an American songwriting and record production team consisting of Bruno Mars, Philip Lawrence and Ari Levine established in a state in the Western United States which is the most populous and the third-largest United State.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Treasure (Bruno Mars song)",
"The Smeezingtons",
"California"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4565,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
74796
|
NADPH dehydrogenase is an enzyme that catalyzes ( increasing the rate of a chemical reaction) the chemical reaction and has two substrates.
|
{
"evidence": [
"NADPH dehydrogenase_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:\nNADPH dehydrogenase is an enzyme that catalyzes ( increasing the rate of a chemical reaction) the chemical reaction and has two substrates.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"NADPH dehydrogenase_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4566,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.071",
"value": "?search="
},
{
"operation": "search",
"time": "11.382",
"value": "NADPH dehydrogenase"
},
{
"operation": "Now on",
"time": "12.246",
"value": "NADPH dehydrogenase"
},
{
"operation": "Highlighting",
"time": "18.723",
"value": "NADPH dehydrogenase_sentence_0"
},
{
"operation": "hyperlink",
"time": "20.762",
"value": "catalyzes"
},
{
"operation": "Now on",
"time": "22.426",
"value": "Catalysis"
},
{
"operation": "Highlighting",
"time": "28.147",
"value": "Catalysis_sentence_3"
},
{
"operation": "Highlighting",
"time": "28.148",
"value": "Catalysis_title"
},
{
"operation": "back-button-clicked",
"time": "30.459",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "30.724",
"value": "NADPH dehydrogenase"
},
{
"operation": "Highlighting",
"time": "38.507",
"value": "NADPH dehydrogenase_sentence_0"
},
{
"operation": "Highlighting",
"time": "39.306",
"value": "NADPH dehydrogenase_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "46.418",
"value": "NADPH dehydrogenase_sentence_0"
},
{
"operation": "finish",
"time": "53.939",
"value": "finish"
}
]
}
|
exfever_6458
|
Billy Bob Thornton is an American actor who starred in a 1992 American crime thriller film directed by an American actor, screenwriter, producer, film, and television director who is a graduate of the University of California, Berkeley.
|
{
"evidence": [
"Billy Bob Thornton",
"One False Move",
"Carl Franklin"
],
"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:\nBilly Bob Thornton is an American actor who starred in a 1992 American crime thriller film directed by an American actor, screenwriter, producer, film, and television director who is a graduate of the University of California, Berkeley.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Billy Bob Thornton",
"One False Move",
"Carl Franklin"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4567,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_415
|
Lumen Ann Pierce is a fictional character portrayed by an Asian actress.
|
{
"evidence": [
"Lumen Pierce",
"Julia Stiles"
],
"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:\nLumen Ann Pierce is a fictional character portrayed by an Asian actress.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Lumen Pierce",
"Julia Stiles"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4568,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_7196
|
Wayne Rooney is an English professional football manager for the highest division of the English Football League, which was previously known as the Division of the Football League in Italy from 1888 until 2004.
|
{
"evidence": [
"Wayne Rooney",
"EFL Championship",
"Football League First Division"
],
"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:\nWayne Rooney is an English professional football manager for the highest division of the English Football League, which was previously known as the Division of the Football League in Italy from 1888 until 2004.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Wayne Rooney",
"EFL Championship",
"Football League First Division"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4569,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
21673
|
Săuzeni, is a 30 km tributary of the river Voinești in Russia used mainly for fish farming.
|
{
"evidence": [
"Săuzeni_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:\nSăuzeni, is a 30 km tributary of the river Voinești in Russia used mainly for fish farming.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Săuzeni_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4570,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.584",
"value": "?search="
},
{
"operation": "search",
"time": "6.689",
"value": "Săuzeni"
},
{
"operation": "Now on",
"time": "7.553",
"value": "Săuzeni"
},
{
"operation": "Highlighting",
"time": "9.803",
"value": "Săuzeni_sentence_2"
},
{
"operation": "Highlighting",
"time": "14.856",
"value": "Săuzeni_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "19.805",
"value": "Săuzeni_sentence_2"
},
{
"operation": "finish",
"time": "39.452",
"value": "finish"
}
]
}
|
37629
|
The village has a total number of 382 houses and the population of 2233 of which 1154 are males while 1179 are females according to the report published by Census India in 2011 and the literacy rate of the village is 77.53%.
|
{
"evidence": [
"2011 Census of India_sentence_8"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe village has a total number of 382 houses and the population of 2233 of which 1154 are males while 1179 are females according to the report published by Census India in 2011 and the literacy rate of the village is 77.53%.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"2011 Census of India_sentence_8"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4571,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.684",
"value": "?search="
},
{
"operation": "search",
"time": "235.479",
"value": "Census India in 2011"
},
{
"operation": "Now on",
"time": "237.531",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "245.66",
"value": "2011 Census of India"
},
{
"operation": "Now on",
"time": "246.561",
"value": "2011 Census of India"
},
{
"operation": "Page search",
"time": "251.983",
"value": "village"
},
{
"operation": "Page search",
"time": "267.103",
"value": "village"
},
{
"operation": "Highlighting",
"time": "273.78",
"value": "2011 Census of India_sentence_8"
},
{
"operation": "finish",
"time": "276.838",
"value": "finish"
}
]
}
|
exfever_7487
|
Grace Jones has been an inspiration for artists, such as the songwriter for Sony/ATV Music Publishing, where she signed a joint deal with a Senegalese American singer, songwriter, record producer, and entrepreneur.
|
{
"evidence": [
"Grace Jones",
"Lady Gaga",
"Akon"
],
"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:\nGrace Jones has been an inspiration for artists, such as the songwriter for Sony/ATV Music Publishing, where she signed a joint deal with a Senegalese American singer, songwriter, record producer, and entrepreneur.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Grace Jones",
"Lady Gaga",
"Akon"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4572,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
5959
|
Souhayr Belhassen was defended human rights and in 2004, joined the International Federation for Human Rights.
|
{
"evidence": [
"Souhayr Belhassen_sentence_0",
"Souhayr Belhassen_sentence_11"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nSouhayr Belhassen was defended human rights and in 2004, joined the International Federation for Human Rights.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Souhayr Belhassen_sentence_0",
"Souhayr Belhassen_sentence_11"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4573,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.479",
"value": "?search="
},
{
"operation": "search",
"time": "16.505",
"value": "Souhayr Belhassen"
},
{
"operation": "Now on",
"time": "18.187",
"value": "Souhayr Belhassen"
},
{
"operation": "Highlighting",
"time": "21.85",
"value": "Souhayr Belhassen_sentence_0"
},
{
"operation": "Highlighting",
"time": "39.166",
"value": "Souhayr Belhassen_sentence_11"
},
{
"operation": "finish",
"time": "80.434",
"value": "finish"
}
]
}
|
exfever_1146
|
Ron Dennis is best known as the founder of a British motor racing team.
|
{
"evidence": [
"Ron Dennis",
"McLaren"
],
"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:\nRon Dennis is best known as the founder of a British motor racing team.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Ron Dennis",
"McLaren"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4574,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3747
|
Channing Tatum is an American actor whose breakout role was as werewolf Alcide Herveaux in the series True Blood, which was shown on an American pay television network launched on November 8, 1972.
|
{
"evidence": [
"Joe Manganiello",
"HBO"
],
"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:\nChanning Tatum is an American actor whose breakout role was as werewolf Alcide Herveaux in the series True Blood, which was shown on an American pay television network launched on November 8, 1972.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Joe Manganiello",
"HBO"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4575,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
18977
|
In 1996, the Short Brothers sold the complete design of Shorland S600 to British Aerospace Australia; then after two years, the Kuwait National Guard ordered 22 S600 in five different versions.
|
{
"evidence": [
"Shorland S600_sentence_2",
"Shorland S600_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:\nIn 1996, the Short Brothers sold the complete design of Shorland S600 to British Aerospace Australia; then after two years, the Kuwait National Guard ordered 22 S600 in five different versions.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Shorland S600_sentence_2",
"Shorland S600_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4576,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.543",
"value": "?search="
},
{
"operation": "search",
"time": "10.441",
"value": "Short Brothers"
},
{
"operation": "Now on",
"time": "11.436",
"value": "Short Brothers"
},
{
"operation": "Page search",
"time": "22.506",
"value": "S600"
},
{
"operation": "Page search",
"time": "23.851",
"value": "S600"
},
{
"operation": "Page search",
"time": "33.562",
"value": "Shorland"
},
{
"operation": "search",
"time": "44.463",
"value": "Shorland S600"
},
{
"operation": "Now on",
"time": "45.444",
"value": "Shorland S600"
},
{
"operation": "Highlighting",
"time": "52.565",
"value": "Shorland S600_sentence_2"
},
{
"operation": "Highlighting",
"time": "74.719",
"value": "Shorland S600_sentence_3"
},
{
"operation": "finish",
"time": "88.317",
"value": "finish"
}
]
}
|
exfever_4393
|
Heroin is used as a recreational drug for its euphoric effects classified as a substance that acts on opioid receptors to produce morphine-like effects.
|
{
"evidence": [
"Heroin",
"Opioid"
],
"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:\nHeroin is used as a recreational drug for its euphoric effects classified as a substance that acts on opioid receptors to produce morphine-like effects.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Heroin",
"Opioid"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4577,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
39607
|
West Virginia Route 480 was called West Virginia Route 48 for a many years.
|
{
"evidence": [
"West Virginia Route 480_sentence_3",
"West Virginia Route 480_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:\nWest Virginia Route 480 was called West Virginia Route 48 for a many years.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"West Virginia Route 480_sentence_3",
"West Virginia Route 480_sentence_6"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4578,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.776",
"value": "?search="
},
{
"operation": "search",
"time": "10.782",
"value": "West Virginia Route 480"
},
{
"operation": "Now on",
"time": "11.789",
"value": "West Virginia Route 480"
},
{
"operation": "Highlighting",
"time": "18.298",
"value": "West Virginia Route 480_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "27.454",
"value": "West Virginia Route 480_sentence_0"
},
{
"operation": "Highlighting",
"time": "48.447",
"value": "West Virginia Route 480_sentence_3"
},
{
"operation": "Highlighting",
"time": "64.783",
"value": "West Virginia Route 480_sentence_6"
},
{
"operation": "finish",
"time": "86.63",
"value": "finish"
}
]
}
|
exfever_8737
|
Reg Watson was an American television producer and screenwriter who created an Australian television soap opera produced by the Grundy Organisation, which was conceived by an Australian television producer and screenwriter.
|
{
"evidence": [
"Reg Watson",
"Prisoner (TV series)",
"Reg Watson"
],
"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:\nReg Watson was an American television producer and screenwriter who created an Australian television soap opera produced by the Grundy Organisation, which was conceived by an Australian television producer and screenwriter.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Reg Watson",
"Prisoner (TV series)",
"Reg Watson"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4579,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_322
|
Wi-Fi was the world's first wireless network.
|
{
"evidence": [
"Wi-Fi",
"Internet"
],
"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:\nWi-Fi was the world's first wireless network.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Wi-Fi",
"Internet"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 4580,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_2776
|
Hogwarts School of Witchcraft and Wizardry is a fictional American boarding school of magic that is the primary setting for a series of seven fantasy novels written by British author J. K. Rowling.
|
{
"evidence": [
"Hogwarts",
"Harry Potter"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHogwarts School of Witchcraft and Wizardry is a fictional American boarding school of magic that is the primary setting for a series of seven fantasy novels written by British author J. K. Rowling.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Hogwarts",
"Harry Potter"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 4581,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
4529
|
Reuben White, who was a tavern owner and a Commissioner, was a member of the 8th, 9th and 11th Legislative Assembly, which is the lower house in the Parliament of of Upper Canada.
|
{
"evidence": [
"Reuben White_sentence_9",
"Reuben White_sentence_8",
"Reuben White_sentence_6",
"Reuben White_sentence_7",
"Legislative Assembly of Upper Canada_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:\nReuben White, who was a tavern owner and a Commissioner, was a member of the 8th, 9th and 11th Legislative Assembly, which is the lower house in the Parliament of of Upper Canada.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Reuben White_sentence_9",
"Reuben White_sentence_8",
"Reuben White_sentence_6",
"Reuben White_sentence_7",
"Legislative Assembly of Upper Canada_sentence_0"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4582,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "14.996",
"value": "Reuben White"
},
{
"operation": "Page search",
"time": "25.878",
"value": "tavern"
},
{
"operation": "Highlighting",
"time": "30.468",
"value": "Reuben White_sentence_9"
},
{
"operation": "Page search",
"time": "40.332",
"value": "commissioner"
},
{
"operation": "Highlighting",
"time": "47.748",
"value": "Reuben White_sentence_8"
},
{
"operation": "Highlighting",
"time": "56.408",
"value": "Reuben White_sentence_6"
},
{
"operation": "Highlighting",
"time": "59.466",
"value": "Reuben White_sentence_7"
},
{
"operation": "hyperlink",
"time": "61.868",
"value": "Legislative Assembly of Upper Canada"
},
{
"operation": "Highlighting",
"time": "68.941",
"value": "Legislative Assembly of Upper Canada_sentence_0"
},
{
"operation": "finish",
"time": "77.803",
"value": "finish"
}
]
}
|
54237
|
With the support of Gertrudis Tenorio Zavala and Cristina Farfán, Rita Cetina Gutiérrez, who wrote poetry from about 1860 which was published in many local journals and newspapers, opened La Siempreviva (Everlasting), an institution consisting of Mexico's first secular school for poor girls as well as an art college for young women.
|
{
"evidence": [
"Rita Cetina Gutiérrez_sentence_8",
"Rita Cetina Gutiérrez_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:\nWith the support of Gertrudis Tenorio Zavala and Cristina Farfán, Rita Cetina Gutiérrez, who wrote poetry from about 1860 which was published in many local journals and newspapers, opened La Siempreviva (Everlasting), an institution consisting of Mexico's first secular school for poor girls as well as an art college for young women.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Rita Cetina Gutiérrez_sentence_8",
"Rita Cetina Gutiérrez_sentence_6"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 4583,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.302",
"value": "?search="
},
{
"operation": "search",
"time": "24.741",
"value": "Rita Cetina Gutiérrez"
},
{
"operation": "Now on",
"time": "25.264",
"value": "Rita Cetina Gutiérrez"
},
{
"operation": "Highlighting",
"time": "34.887",
"value": "Rita Cetina Gutiérrez_sentence_8"
},
{
"operation": "Highlighting",
"time": "53.595",
"value": "Rita Cetina Gutiérrez_sentence_6"
},
{
"operation": "finish",
"time": "134.746",
"value": "finish"
}
]
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.