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_5546
|
Stern became one of the highest-paid radio figures worth $500 million.
|
{
"evidence": [
"Howard Stern",
"Detroit"
],
"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:\nStern became one of the highest-paid radio figures worth $500 million.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Howard Stern",
"Detroit"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6584,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_2348
|
Cleopatra was Queen of the Ptolemaic Kingdom of Israel, a descendant of Ptolemy I Soter, who came from the ancient tribe that lived on the longest river that flows fully into Greece.
|
{
"evidence": [
"Cleopatra",
"Ancient Macedonians",
"Haliacmon"
],
"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 was Queen of the Ptolemaic Kingdom of Israel, a descendant of Ptolemy I Soter, who came from the ancient tribe that lived on the longest river that flows fully into Greece.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Cleopatra",
"Ancient Macedonians",
"Haliacmon"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6585,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
18148
|
The plot of the silent film (or a film without synchronized recorded sound), The Family Honor, did not tackle the lives of the Tucker family.
|
{
"evidence": [
"The Family Honor_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 plot of the silent film (or a film without synchronized recorded sound), The Family Honor, did not tackle the lives of the Tucker family.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"The Family Honor_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6586,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.61",
"value": "?search="
},
{
"operation": "search",
"time": "19.58",
"value": "The Family Honor"
},
{
"operation": "Now on",
"time": "20.298",
"value": "The Family Honor"
},
{
"operation": "hyperlink",
"time": "34.216",
"value": "The Family Honor (1917 film)"
},
{
"operation": "Now on",
"time": "34.835",
"value": "The Family Honor (1917 film)"
},
{
"operation": "back-button-clicked",
"time": "43.233",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "43.33",
"value": "The Family Honor"
},
{
"operation": "Highlighting",
"time": "51.039",
"value": "The Family Honor_sentence_1"
},
{
"operation": "Highlighting",
"time": "58.863",
"value": "The Family Honor_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "66.102",
"value": "The Family Honor_sentence_1"
},
{
"operation": "finish",
"time": "73.204",
"value": "finish"
}
]
}
|
exfever_5581
|
Harold Allen Ramis was an American actor, best known for his role as Egon Spengler in a 1989 American supernatural comedy film directed and produced by Dan Aykroyd.
|
{
"evidence": [
"Harold Ramis",
"Ghostbusters II"
],
"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:\nHarold Allen Ramis was an American actor, best known for his role as Egon Spengler in a 1989 American supernatural comedy film directed and produced by Dan Aykroyd.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Harold Ramis",
"Ghostbusters II"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6587,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_4973
|
The Golden Girls is an American movie about four older women who share a home in a state located in the Southeastern region bordered by the 24th-largest area in the U.S.
|
{
"evidence": [
"The Golden Girls",
"Florida",
"Georgia (U.S. state)"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Golden Girls is an American movie about four older women who share a home in a state located in the Southeastern region bordered by the 24th-largest area in the U.S.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Golden Girls",
"Florida",
"Georgia (U.S. state)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6588,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5377
|
The eighth season of an American singing competition television series created by Simon Fuller premiered on Tuesday, January 13, 2009.
|
{
"evidence": [
"American Idol (season 8)",
"American Idol"
],
"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 eighth season of an American singing competition television series created by Simon Fuller premiered on Tuesday, January 13, 2009.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"American Idol (season 8)",
"American Idol"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6589,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
76162
|
Nkabika worked for the Zambia National Provident Fund and ran the NODA Investment company, then he contested the Kapiri Mposhi seat as an independent candidate in the 1996 general elections.
|
{
"evidence": [
"Macdonald Nkabika_sentence_2",
"Macdonald Nkabika_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:\nNkabika worked for the Zambia National Provident Fund and ran the NODA Investment company, then he contested the Kapiri Mposhi seat as an independent candidate in the 1996 general elections.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Macdonald Nkabika_sentence_2",
"Macdonald Nkabika_sentence_3"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6590,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.473",
"value": "?search="
},
{
"operation": "search",
"time": "12.83",
"value": "Contains...Nkabika"
},
{
"operation": "Now on",
"time": "13.645",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "18.049",
"value": "Macdonald Nkabika"
},
{
"operation": "Now on",
"time": "18.582",
"value": "Macdonald Nkabika"
},
{
"operation": "Page search",
"time": "27.538",
"value": "provident"
},
{
"operation": "Highlighting",
"time": "30.43",
"value": "Macdonald Nkabika_sentence_2"
},
{
"operation": "Highlighting",
"time": "44.775",
"value": "Macdonald Nkabika_sentence_3"
},
{
"operation": "finish",
"time": "261.698",
"value": "finish"
}
]
}
|
exfever_2842
|
Ronin is a 1998 American action thriller film written by John David Zeik and an American playwright, filmmaker, and author who won a Pulitzer Prize.
|
{
"evidence": [
"Ronin (film)",
"David Mamet"
],
"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:\nRonin is a 1998 American action thriller film written by John David Zeik and an American playwright, filmmaker, and author who won a Pulitzer Prize.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Ronin (film)",
"David Mamet"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6591,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
69460
|
Yingshouyingzi Mining District belongs to the prefecture-level city of Chengde in the province of Hebei, China.
|
{
"evidence": [
"Yingshouyingzi Mining District_sentence_0",
"Chengde_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:\nYingshouyingzi Mining District belongs to the prefecture-level city of Chengde in the province of Hebei, China.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Yingshouyingzi Mining District_sentence_0",
"Chengde_sentence_2"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6592,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.379",
"value": "?search="
},
{
"operation": "search",
"time": "11.1",
"value": "Yingshouyingzi Mining District"
},
{
"operation": "Now on",
"time": "13.091",
"value": "Yingshouyingzi Mining District"
},
{
"operation": "Highlighting",
"time": "22.368",
"value": "Yingshouyingzi Mining District_sentence_0"
},
{
"operation": "hyperlink",
"time": "23.751",
"value": "Chengde"
},
{
"operation": "Now on",
"time": "28.412",
"value": "Chengde"
},
{
"operation": "Highlighting",
"time": "38.695",
"value": "Chengde_sentence_2"
},
{
"operation": "finish",
"time": "49.455",
"value": "finish"
}
]
}
|
exfever_4768
|
Andy Warhol's works explore the relationship between artistic expression, celebrity culture and a marketing communication that employs an openly sponsored, non-personal message to promote or sell a product, service or idea.
|
{
"evidence": [
"Andy Warhol",
"Advertising"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAndy Warhol's works explore the relationship between artistic expression, celebrity culture and a marketing communication that employs an openly sponsored, non-personal message to promote or sell a product, service or idea.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Andy Warhol",
"Advertising"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6593,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1577
|
Resident Evil is a 2002 sci-fi action horror film written and directed by Paul W. S. Anderson that stars a British actor who played General Zod in an American television series by David S. Goyer.
|
{
"evidence": [
"Resident Evil (film)",
"Colin Salmon",
"Krypton (TV series)"
],
"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:\nResident Evil is a 2002 sci-fi action horror film written and directed by Paul W. S. Anderson that stars a British actor who played General Zod in an American television series by David S. Goyer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Resident Evil (film)",
"Colin Salmon",
"Krypton (TV series)"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6594,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_2398
|
King Kong is a 2005 adventure monster film directed by Peter Jackson that features a large-scale, sweeping scope, and spectacle.
|
{
"evidence": [
"King Kong (2005 film)",
"Epic film"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nKing Kong is a 2005 adventure monster film directed by Peter Jackson that features a large-scale, sweeping scope, and spectacle.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"King Kong (2005 film)",
"Epic film"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6595,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1916
|
Colin Farrell is the youngest Irish actor to have appeared in the BBC drama series Ballykissangel.
|
{
"evidence": [
"Colin Farrell",
"S.W.A.T. (film)",
"Thriller film"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nColin Farrell is the youngest Irish actor to have appeared in the BBC drama series Ballykissangel.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Colin Farrell",
"S.W.A.T. (film)",
"Thriller film"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6596,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
90788
|
William Wallace Denslow was born in Philadelphia and spent some time at the National Academy of Design (an honorary association of American artists, founded in New York City), and the 1880's he traveled about the United States as an artist and newspaper reporter.
|
{
"evidence": [
"William Wallace Denslow_sentence_3",
"National Academy of Design_sentence_0",
"William Wallace Denslow_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:\nWilliam Wallace Denslow was born in Philadelphia and spent some time at the National Academy of Design (an honorary association of American artists, founded in New York City), and the 1880's he traveled about the United States as an artist and newspaper reporter.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"William Wallace Denslow_sentence_3",
"National Academy of Design_sentence_0",
"William Wallace Denslow_sentence_4"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6597,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.414",
"value": "?search="
},
{
"operation": "search",
"time": "21.364",
"value": "William Wallace Denslow"
},
{
"operation": "Now on",
"time": "22.282",
"value": "William Wallace Denslow"
},
{
"operation": "Highlighting",
"time": "85.736",
"value": "William Wallace Denslow_sentence_3"
},
{
"operation": "hyperlink",
"time": "89.975",
"value": "National Academy of Design"
},
{
"operation": "Now on",
"time": "90.844",
"value": "National Academy Museum and School"
},
{
"operation": "Highlighting",
"time": "92.684",
"value": "National Academy of Design_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "95.406",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "95.503",
"value": "William Wallace Denslow"
},
{
"operation": "Highlighting",
"time": "96.496",
"value": "William Wallace Denslow_sentence_4"
},
{
"operation": "finish",
"time": "103.435",
"value": "finish"
}
]
}
|
18663
|
Aitken is a player who played for the St. George Illawarra Dragons, played for the Australian Schoolboys and named to play for the New South Wales.
|
{
"evidence": [
"Euan Aitken_sentence_8",
"Euan Aitken_sentence_9",
"Euan Aitken_sentence_10"
],
"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:\nAitken is a player who played for the St. George Illawarra Dragons, played for the Australian Schoolboys and named to play for the New South Wales.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"Euan Aitken_sentence_8",
"Euan Aitken_sentence_9",
"Euan Aitken_sentence_10"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6598,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.582",
"value": "?search="
},
{
"operation": "search",
"time": "6.255",
"value": "Aitken"
},
{
"operation": "Now on",
"time": "6.916",
"value": "Aitken"
},
{
"operation": "search",
"time": "32.13",
"value": "Euan Aitken"
},
{
"operation": "Now on",
"time": "32.994",
"value": "Euan Aitken"
},
{
"operation": "Page search",
"time": "42.692",
"value": "George"
},
{
"operation": "Highlighting",
"time": "50.896",
"value": "Euan Aitken_sentence_8"
},
{
"operation": "Highlighting",
"time": "62.666",
"value": "Euan Aitken_sentence_9"
},
{
"operation": "Highlighting",
"time": "71.897",
"value": "Euan Aitken_sentence_10"
},
{
"operation": "finish",
"time": "78.668",
"value": "finish"
}
]
}
|
exfever_8084
|
The Council on Foreign Relations is a legal entity organized and operated for a collective, public or social benefit, which may seek approval from the government to remove liability to make a compulsory payment that would otherwise be imposed by a ruling power.
|
{
"evidence": [
"Council on Foreign Relations",
"Nonprofit organization",
"Tax exemption"
],
"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 Council on Foreign Relations is a legal entity organized and operated for a collective, public or social benefit, which may seek approval from the government to remove liability to make a compulsory payment that would otherwise be imposed by a ruling power.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Council on Foreign Relations",
"Nonprofit organization",
"Tax exemption"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6599,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5951
|
The Joker, created by Bill Finger, is a supervillain appearing in Japanese comic books published by an American comic book publisher and the flagship unit of DC Entertainment, a subsidiary of Warner Bros. Discovery.
|
{
"evidence": [
"Joker (character)",
"DC Comics"
],
"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 Joker, created by Bill Finger, is a supervillain appearing in Japanese comic books published by an American comic book publisher and the flagship unit of DC Entertainment, a subsidiary of Warner Bros. Discovery.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Joker (character)",
"DC Comics"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6600,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
33648
|
The 2010 Gascoyne River flood, triggered by record-breaking rainfall of over 6,000% of monthly mean, has severely affected the coastal town of Carnarvon that has a population of 4,426.
|
{
"evidence": [
"2010 Gascoyne River flood_sentence_1",
"2010 Gascoyne River flood_sentence_5",
"Carnarvon, Western Australia_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:\nThe 2010 Gascoyne River flood, triggered by record-breaking rainfall of over 6,000% of monthly mean, has severely affected the coastal town of Carnarvon that has a population of 4,426.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"2010 Gascoyne River flood_sentence_1",
"2010 Gascoyne River flood_sentence_5",
"Carnarvon, Western Australia_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6601,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.917",
"value": "?search="
},
{
"operation": "search",
"time": "4.827",
"value": "2010 Gascoyne River flood"
},
{
"operation": "Now on",
"time": "5.49",
"value": "2010 Gascoyne River flood"
},
{
"operation": "Highlighting",
"time": "11.822",
"value": "2010 Gascoyne River flood_sentence_1"
},
{
"operation": "Page search",
"time": "29.486",
"value": "Carnarvon"
},
{
"operation": "Highlighting",
"time": "31.238",
"value": "2010 Gascoyne River flood_sentence_5"
},
{
"operation": "hyperlink",
"time": "35.907",
"value": "Carnarvon"
},
{
"operation": "Now on",
"time": "36.422",
"value": "Carnarvon, Western Australia"
},
{
"operation": "Highlighting",
"time": "38.087",
"value": "Carnarvon, Western Australia_sentence_5"
},
{
"operation": "finish",
"time": "333.681",
"value": "finish"
}
]
}
|
85111
|
Rudolf Ryban was elected to the Slovak Student Council from 1935-1937, and also elected and became Chair of the Hurban Club at the Faculty of Theology and Chair of Dr. Michal Lucansky's National Lutheran Youth Organization.
|
{
"evidence": [
"Ondrej Bartko_sentence_8",
"Ondrej Bartko_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:\nRudolf Ryban was elected to the Slovak Student Council from 1935-1937, and also elected and became Chair of the Hurban Club at the Faculty of Theology and Chair of Dr. Michal Lucansky's National Lutheran Youth Organization.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Ondrej Bartko_sentence_8",
"Ondrej Bartko_sentence_10"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6602,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.984",
"value": "?search="
},
{
"operation": "search",
"time": "40.628",
"value": "rudolf ryban"
},
{
"operation": "Now on",
"time": "41.509",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "46.212",
"value": "Ondrej Bartko"
},
{
"operation": "Now on",
"time": "47.019",
"value": "Ondrej Bartko"
},
{
"operation": "Page search",
"time": "50.844",
"value": "ryban"
},
{
"operation": "search",
"time": "84.285",
"value": "rudolf ryban slovak"
},
{
"operation": "Now on",
"time": "85.346",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "89.045",
"value": "rudolf rybak slovak"
},
{
"operation": "Now on",
"time": "89.933",
"value": "undefined"
},
{
"operation": "search",
"time": "102.732",
"value": "ryban"
},
{
"operation": "Now on",
"time": "103.765",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "106.372",
"value": "rybak"
},
{
"operation": "Now on",
"time": "107.254",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "113.893",
"value": "Rybak"
},
{
"operation": "Now on",
"time": "116.082",
"value": "Rybak"
},
{
"operation": "search",
"time": "143.373",
"value": "slovak student council"
},
{
"operation": "Now on",
"time": "144.579",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "160.437",
"value": "Ondrej Bartko"
},
{
"operation": "Now on",
"time": "160.754",
"value": "Ondrej Bartko"
},
{
"operation": "Page search",
"time": "170.093",
"value": "ryban"
},
{
"operation": "Page search",
"time": "202.332",
"value": "student"
},
{
"operation": "Page search",
"time": "229.325",
"value": "hurban"
},
{
"operation": "Highlighting",
"time": "236.669",
"value": "Ondrej Bartko_sentence_10"
},
{
"operation": "Highlighting",
"time": "247.604",
"value": "Ondrej Bartko_sentence_8"
},
{
"operation": "Page search",
"time": "264.124",
"value": "chair"
},
{
"operation": "Highlighting deleted",
"time": "269.548",
"value": "Ondrej Bartko_sentence_10"
},
{
"operation": "Highlighting",
"time": "270.604",
"value": "Ondrej Bartko_sentence_10"
},
{
"operation": "finish",
"time": "280.94",
"value": "finish"
}
]
}
|
exfever_7871
|
Gordon James Ramsay is a British chef who became one of the best-known and most influential chefs in a sovereign country in north-western Europe that have been separated by the sea of people who spoke a Goidelic language of the Insular Celtic branch of the Celtic language family, which is a part of the Indo-European language family.
|
{
"evidence": [
"Gordon Ramsay",
"United Kingdom",
"Irish language"
],
"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:\nGordon James Ramsay is a British chef who became one of the best-known and most influential chefs in a sovereign country in north-western Europe that have been separated by the sea of people who spoke a Goidelic language of the Insular Celtic branch of the Celtic language family, which is a part of the Indo-European language family.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Gordon Ramsay",
"United Kingdom",
"Irish language"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6603,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
26832
|
Jinshi City is in the province of Hunan and is divided into six towns.
|
{
"evidence": [
"Jinshi City_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:\nJinshi City is in the province of Hunan and is divided into six towns.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Jinshi City_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6604,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.395",
"value": "?search="
},
{
"operation": "search",
"time": "7.169",
"value": "Jinshi City"
},
{
"operation": "Now on",
"time": "9.239",
"value": "Jinshi City"
},
{
"operation": "Highlighting",
"time": "18.521",
"value": "Jinshi City_cell_0_2_1"
},
{
"operation": "Highlighting",
"time": "18.543",
"value": "Jinshi City_header_cell_0_2_0"
},
{
"operation": "Highlighting",
"time": "18.545",
"value": "Jinshi City_header_cell_0_0_0"
},
{
"operation": "Highlighting deleted",
"time": "18.875",
"value": "Jinshi City_cell_0_2_1"
},
{
"operation": "Highlighting deleted",
"time": "18.898",
"value": "Jinshi City_header_cell_0_2_0"
},
{
"operation": "Highlighting deleted",
"time": "18.902",
"value": "Jinshi City_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "19.449",
"value": "Jinshi City_cell_0_2_1"
},
{
"operation": "Highlighting",
"time": "19.47",
"value": "Jinshi City_header_cell_0_2_0"
},
{
"operation": "Highlighting",
"time": "19.471",
"value": "Jinshi City_header_cell_0_0_0"
},
{
"operation": "Highlighting deleted",
"time": "31.195",
"value": "Jinshi City_cell_0_2_1"
},
{
"operation": "Highlighting deleted",
"time": "31.217",
"value": "Jinshi City_header_cell_0_2_0"
},
{
"operation": "Highlighting deleted",
"time": "31.221",
"value": "Jinshi City_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "32.742",
"value": "Jinshi City_sentence_2"
},
{
"operation": "finish",
"time": "51.281",
"value": "finish"
}
]
}
|
exfever_3754
|
Sir Timothy Miles Bindon Rice is an English lyricist and author who wrote a sung-through musical comedy that is based on the character of Joseph from the Bible's Book of Genesis.
|
{
"evidence": [
"Tim Rice",
"Joseph and the Amazing Technicolor Dreamcoat"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nSir Timothy Miles Bindon Rice is an English lyricist and author who wrote a sung-through musical comedy that is based on the character of Joseph from the Bible's Book of Genesis.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tim Rice",
"Joseph and the Amazing Technicolor Dreamcoat"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6605,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
32203
|
In 2012, Art Agrawal and Dongyi Liao founded Zumper, and in September of the same year, was able to raise $1.8 million in seed funding from venture firms and investors, and also won the TechCrunch Battlefield Disrupt San Francisco competition.
|
{
"evidence": [
"YourMechanic_sentence_8",
"YourMechanic_sentence_9",
"YourMechanic_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:\nIn 2012, Art Agrawal and Dongyi Liao founded Zumper, and in September of the same year, was able to raise $1.8 million in seed funding from venture firms and investors, and also won the TechCrunch Battlefield Disrupt San Francisco competition.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"YourMechanic_sentence_8",
"YourMechanic_sentence_9",
"YourMechanic_sentence_10"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6606,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.395",
"value": "?search="
},
{
"operation": "search",
"time": "18.014",
"value": "Zumper language"
},
{
"operation": "Now on",
"time": "19.136",
"value": "Kuteb language"
},
{
"operation": "search",
"time": "31.515",
"value": "Contains...Zumper"
},
{
"operation": "Now on",
"time": "32.343",
"value": "undefined"
},
{
"operation": "search",
"time": "389.365",
"value": "yourmechanic"
},
{
"operation": "Now on",
"time": "390.571",
"value": "YourMechanic"
},
{
"operation": "Highlighting",
"time": "426.953",
"value": "YourMechanic_sentence_8"
},
{
"operation": "Highlighting",
"time": "430.56",
"value": "YourMechanic_sentence_9"
},
{
"operation": "Highlighting",
"time": "435.22",
"value": "YourMechanic_sentence_10"
},
{
"operation": "finish",
"time": "453.016",
"value": "finish"
}
]
}
|
exfever_861
|
Broadcast News is an American romantic comedy-drama film starring an American actress and fashion model known for her role in an American romantic comedy-drama film written by James L. Brooks.
|
{
"evidence": [
"Broadcast News (film)",
"Lois Chiles",
"Broadcast News (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:\nBroadcast News is an American romantic comedy-drama film starring an American actress and fashion model known for her role in an American romantic comedy-drama film written by James L. Brooks.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Broadcast News (film)",
"Lois Chiles",
"Broadcast News (film)"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6607,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
30048
|
The coach of the 2009–10 FC Barcelona season, Pep Guardiola, who joined the team La Masia at the age of 16, claimed that the team is united and dedicated and the players remain focused despite the scandal.
|
{
"evidence": [
"Pep Guardiola_sentence_19"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe coach of the 2009–10 FC Barcelona season, Pep Guardiola, who joined the team La Masia at the age of 16, claimed that the team is united and dedicated and the players remain focused despite the scandal.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Pep Guardiola_sentence_19"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6608,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.871",
"value": "undefined"
},
{
"operation": "search",
"time": "12.081",
"value": "Pep Guardiola"
},
{
"operation": "Now on",
"time": "13.689",
"value": "Pep Guardiola"
},
{
"operation": "Page search",
"time": "19.284",
"value": "maisa"
},
{
"operation": "Page search",
"time": "38.351",
"value": "masia"
},
{
"operation": "Highlighting",
"time": "45.069",
"value": "Pep Guardiola_sentence_19"
},
{
"operation": "hyperlink",
"time": "63.954",
"value": "Romário"
},
{
"operation": "Now on",
"time": "65.351",
"value": "Romário"
},
{
"operation": "back-button-clicked",
"time": "169.707",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "169.971",
"value": "Pep Guardiola"
},
{
"operation": "finish",
"time": "180.293",
"value": "finish"
}
]
}
|
exfever_8862
|
American Pie is a 1999 American coming-of-age sex comedy film written by an American screenwriter and producer.
|
{
"evidence": [
"American Pie (film)",
"Adam Herz"
],
"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:\nAmerican Pie is a 1999 American coming-of-age sex comedy film written by an American screenwriter and producer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"American Pie (film)",
"Adam Herz"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6609,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
60802
|
The flag of Naples had a castle on it to honor its former kingdom.
|
{
"evidence": [
"Flag of Naples_sentence_1"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe flag of Naples had a castle on it to honor its former kingdom.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Flag of Naples_sentence_1"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6610,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.208",
"value": "?search="
},
{
"operation": "search",
"time": "6.134",
"value": "flag of Naples"
},
{
"operation": "Now on",
"time": "8.428",
"value": "Flag of Naples"
},
{
"operation": "Highlighting",
"time": "32.404",
"value": "Flag of Naples_sentence_1"
},
{
"operation": "finish",
"time": "94.03",
"value": "finish"
}
]
}
|
16859
|
Zionism is an organized movement founded in 1897 by Theodor Herzl and is related to Judaism and Jewish history. in which the Bund which was founded in 1897 became one of the largest socialist movements in Europe and campaigned for Jewish autonomy and recognition national rights within Russia.
|
{
"evidence": [
"History of Zionism_sentence_200",
"History of Zionism_sentence_199",
"History of Zionism_sentence_198",
"History of Zionism_sentence_0",
"History of Zionism_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:\nZionism is an organized movement founded in 1897 by Theodor Herzl and is related to Judaism and Jewish history. in which the Bund which was founded in 1897 became one of the largest socialist movements in Europe and campaigned for Jewish autonomy and recognition national rights within Russia.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"History of Zionism_sentence_200",
"History of Zionism_sentence_199",
"History of Zionism_sentence_198",
"History of Zionism_sentence_0",
"History of Zionism_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6611,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.631",
"value": "?search="
},
{
"operation": "search",
"time": "1142.152",
"value": "Zionism"
},
{
"operation": "Now on",
"time": "1143.485",
"value": "Zionism"
},
{
"operation": "Page search",
"time": "1157.391",
"value": "1897"
},
{
"operation": "Highlighting",
"time": "1168.293",
"value": "Zionism_sentence_21"
},
{
"operation": "Highlighting",
"time": "1168.294",
"value": "Zionism_title"
},
{
"operation": "Highlighting",
"time": "1168.294",
"value": "Zionism_section_1"
},
{
"operation": "Page search",
"time": "1188.176",
"value": "bund"
},
{
"operation": "search",
"time": "1227.211",
"value": "Bund"
},
{
"operation": "Now on",
"time": "1227.914",
"value": "Bund"
},
{
"operation": "Page search",
"time": "1236.953",
"value": "1897"
},
{
"operation": "Highlighting",
"time": "1244.151",
"value": "Bund_item_4_19"
},
{
"operation": "Highlighting",
"time": "1244.152",
"value": "Bund_title"
},
{
"operation": "Highlighting",
"time": "1244.152",
"value": "Bund_section_4"
},
{
"operation": "Page search",
"time": "1257.901",
"value": "social movement"
},
{
"operation": "Page search",
"time": "1291.095",
"value": "socialist movement"
},
{
"operation": "hyperlink",
"time": "1298.559",
"value": "Bundism"
},
{
"operation": "Now on",
"time": "1299.1",
"value": "Bundism"
},
{
"operation": "Highlighting",
"time": "1307.549",
"value": "Bundism_sentence_0"
},
{
"operation": "Highlighting",
"time": "1307.55",
"value": "Bundism_title"
},
{
"operation": "Page search",
"time": "1323.927",
"value": "socialist movement"
},
{
"operation": "Page search",
"time": "1338",
"value": "jewish autonomy"
},
{
"operation": "Page search",
"time": "1373.022",
"value": "national rights"
},
{
"operation": "back-button-clicked",
"time": "1403.523",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "1403.607",
"value": "Bund"
},
{
"operation": "back-button-clicked",
"time": "1404.356",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "1404.361",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "1404.548",
"value": "Zionism"
},
{
"operation": "Page search",
"time": "1415.441",
"value": "jewish history"
},
{
"operation": "search",
"time": "1459.485",
"value": "campaigned for Jewish autonomy"
},
{
"operation": "Now on",
"time": "1461.046",
"value": "undefined"
},
{
"operation": "search",
"time": "1494.483",
"value": "\"campaigned for Jewish autonomy and recognition\""
},
{
"operation": "Now on",
"time": "1495.085",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "1500.114",
"value": "History of Zionism"
},
{
"operation": "Now on",
"time": "1501.274",
"value": "History of Zionism"
},
{
"operation": "Highlighting",
"time": "1505.774",
"value": "History of Zionism_sentence_0"
},
{
"operation": "Highlighting",
"time": "1511.232",
"value": "History of Zionism_sentence_1"
},
{
"operation": "Page search",
"time": "1528.734",
"value": "jewish autonomy"
},
{
"operation": "Highlighting",
"time": "1551.046",
"value": "History of Zionism_sentence_200"
},
{
"operation": "Highlighting",
"time": "1551.556",
"value": "History of Zionism_sentence_199"
},
{
"operation": "Highlighting",
"time": "1552.296",
"value": "History of Zionism_sentence_198"
},
{
"operation": "Highlighting",
"time": "1562.486",
"value": "History of Zionism_sentence_0"
},
{
"operation": "Highlighting",
"time": "1564.278",
"value": "History of Zionism_sentence_1"
},
{
"operation": "finish",
"time": "1570.808",
"value": "finish"
}
]
}
|
68730
|
Plunket shark, also known as S. plunketi, is part of the scientific classification scymnodon under the somniosidae family which are commonly known as hammerhead sharks..
|
{
"evidence": [
"Somniosidae_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:\nPlunket shark, also known as S. plunketi, is part of the scientific classification scymnodon under the somniosidae family which are commonly known as hammerhead sharks..\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Somniosidae_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6612,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.49",
"value": "?search="
},
{
"operation": "search",
"time": "24.022",
"value": "Plunket shark"
},
{
"operation": "Now on",
"time": "24.935",
"value": "Plunket shark"
},
{
"operation": "Highlighting",
"time": "57.883",
"value": "Plunket shark_cell_0_9_0"
},
{
"operation": "Highlighting",
"time": "57.884",
"value": "Plunket shark_title"
},
{
"operation": "Highlighting",
"time": "57.887",
"value": "Plunket shark_header_cell_0_2_0"
},
{
"operation": "Highlighting",
"time": "57.887",
"value": "Plunket shark_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "57.888",
"value": "Plunket shark_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "58.65",
"value": "Plunket shark_cell_0_9_1"
},
{
"operation": "Highlighting",
"time": "68.41",
"value": "Plunket shark_cell_0_8_1"
},
{
"operation": "Highlighting",
"time": "69.179",
"value": "Plunket shark_cell_0_8_0"
},
{
"operation": "Highlighting deleted",
"time": "74.17",
"value": "Plunket shark_cell_0_8_1"
},
{
"operation": "Highlighting deleted",
"time": "74.762",
"value": "Plunket shark_cell_0_8_0"
},
{
"operation": "Highlighting",
"time": "82.774",
"value": "Plunket shark_cell_0_7_0"
},
{
"operation": "Highlighting",
"time": "83.734",
"value": "Plunket shark_cell_0_7_1"
},
{
"operation": "hyperlink",
"time": "90.578",
"value": "Somniosidae"
},
{
"operation": "Now on",
"time": "91.656",
"value": "Somniosidae"
},
{
"operation": "Highlighting",
"time": "95.199",
"value": "Somniosidae_sentence_0"
},
{
"operation": "finish",
"time": "122.642",
"value": "finish"
}
]
}
|
19347
|
Bing Davis, who holds a position in the National Conference of Artists, accounts clay as is his most expressive medium of art and believes that there's a disconnect when a tool separates him from the work such as a brush or a lens.
|
{
"evidence": [
"Bing Davis_sentence_7",
"Bing Davis_sentence_9",
"Bing Davis_sentence_10"
],
"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:\nBing Davis, who holds a position in the National Conference of Artists, accounts clay as is his most expressive medium of art and believes that there's a disconnect when a tool separates him from the work such as a brush or a lens.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Bing Davis_sentence_7",
"Bing Davis_sentence_9",
"Bing Davis_sentence_10"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6613,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.583",
"value": "?search="
},
{
"operation": "search",
"time": "4.957",
"value": "Bing Davis"
},
{
"operation": "Now on",
"time": "5.613",
"value": "Bing Davis"
},
{
"operation": "Page search",
"time": "14.337",
"value": "National Conference"
},
{
"operation": "Highlighting",
"time": "16.371",
"value": "Bing Davis_sentence_7"
},
{
"operation": "Page search",
"time": "38.679",
"value": "clay"
},
{
"operation": "Highlighting",
"time": "49.988",
"value": "Bing Davis_sentence_9"
},
{
"operation": "Highlighting",
"time": "61.02",
"value": "Bing Davis_sentence_10"
},
{
"operation": "finish",
"time": "71.072",
"value": "finish"
}
]
}
|
53757
|
On 27 January 1969, Iraqi authorities hanged 14 alleged spies (nine Jews, three Muslims and two Christians) for Israel in a public execution in Baghdad where citizens were invited by Baghdad Radio to Liberation Square, which is Baghdad's biggest and most central square, to "come and enjoy the feast".
|
{
"evidence": [
"1969 Baghdad hangings_sentence_0",
"1969 Baghdad hangings_sentence_7",
"Liberation Square, Baghdad_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:\nOn 27 January 1969, Iraqi authorities hanged 14 alleged spies (nine Jews, three Muslims and two Christians) for Israel in a public execution in Baghdad where citizens were invited by Baghdad Radio to Liberation Square, which is Baghdad's biggest and most central square, to \"come and enjoy the feast\".\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"1969 Baghdad hangings_sentence_0",
"1969 Baghdad hangings_sentence_7",
"Liberation Square, Baghdad_sentence_1"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6614,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.687",
"value": "?search="
},
{
"operation": "search",
"time": "24.925",
"value": "Contains...Iraqi authorities hanged 14 alleged spies"
},
{
"operation": "Now on",
"time": "27.744",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "31.097",
"value": "1969 Baghdad hangings"
},
{
"operation": "Now on",
"time": "31.455",
"value": "1969 Baghdad hangings"
},
{
"operation": "Highlighting",
"time": "44.334",
"value": "1969 Baghdad hangings_sentence_0"
},
{
"operation": "Highlighting",
"time": "56.593",
"value": "1969 Baghdad hangings_sentence_7"
},
{
"operation": "hyperlink",
"time": "71.928",
"value": "Liberation Square"
},
{
"operation": "Now on",
"time": "72.652",
"value": "Liberation Square, Baghdad"
},
{
"operation": "Highlighting",
"time": "75.568",
"value": "Liberation Square, Baghdad_sentence_1"
},
{
"operation": "finish",
"time": "105.5",
"value": "finish"
}
]
}
|
56688
|
The role of Rebecca in I Think I Love My Wife, a 2007 American romantic comedy film starring Kerry Washington, Chris Rock and Gina Torres, was played by Kimberly Hébert Gregory.
|
{
"evidence": [
"I Think I Love My Wife_sentence_0",
"Kimberly Hébert Gregory_sentence_0",
"Kimberly Hébert Gregory_sentence_10"
],
"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 role of Rebecca in I Think I Love My Wife, a 2007 American romantic comedy film starring Kerry Washington, Chris Rock and Gina Torres, was played by Kimberly Hébert Gregory.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"I Think I Love My Wife_sentence_0",
"Kimberly Hébert Gregory_sentence_0",
"Kimberly Hébert Gregory_sentence_10"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6615,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.56",
"value": "?search="
},
{
"operation": "search",
"time": "19.04",
"value": "I Think I Love My Wife"
},
{
"operation": "Now on",
"time": "20.646",
"value": "I Think I Love My Wife"
},
{
"operation": "Highlighting",
"time": "24.91",
"value": "I Think I Love My Wife_sentence_0"
},
{
"operation": "search",
"time": "115.457",
"value": "Contains...Kimberly Hébert Gregory."
},
{
"operation": "Now on",
"time": "118.574",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "121.797",
"value": "Kimberly Hébert Gregory"
},
{
"operation": "Now on",
"time": "123.883",
"value": "Kimberly Hébert Gregory"
},
{
"operation": "Highlighting",
"time": "125.575",
"value": "Kimberly Hébert Gregory_sentence_0"
},
{
"operation": "Highlighting",
"time": "141.316",
"value": "Kimberly Hébert Gregory_sentence_10"
},
{
"operation": "hyperlink",
"time": "149.979",
"value": "I Think I Love My Wife"
},
{
"operation": "Now on",
"time": "151.033",
"value": "I Think I Love My Wife"
},
{
"operation": "finish",
"time": "209.072",
"value": "finish"
}
]
}
|
exfever_7527
|
Steve Buscemi is known for his role in Reservoir Dogs, directed by a popular director known for his films that are characterized by frequent references to popular culture.
|
{
"evidence": [
"Steve Buscemi",
"Quentin Tarantino"
],
"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:\nSteve Buscemi is known for his role in Reservoir Dogs, directed by a popular director known for his films that are characterized by frequent references to popular culture.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Steve Buscemi",
"Quentin Tarantino"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6616,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
32360
|
KXMX, originally known as KQCR, changed its format to country music on September 5, 1995.
|
{
"evidence": [
"KZIA_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:\nKXMX, originally known as KQCR, changed its format to country music on September 5, 1995.\n",
"role": "user"
}
] |
Entity Disambiguation
|
{
"ground_truth": {
"evidence": [
"KZIA_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6617,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.731",
"value": "?search="
},
{
"operation": "search",
"time": "9.181",
"value": "KXMX"
},
{
"operation": "Now on",
"time": "10.078",
"value": "KXMX"
},
{
"operation": "hyperlink",
"time": "49.711",
"value": "KGBN"
},
{
"operation": "Now on",
"time": "50.392",
"value": "KGBN"
},
{
"operation": "back-button-clicked",
"time": "64.988",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "65.066",
"value": "KXMX"
},
{
"operation": "search",
"time": "75.188",
"value": "KQCR"
},
{
"operation": "Now on",
"time": "75.923",
"value": "KQCR"
},
{
"operation": "Highlighting",
"time": "85.586",
"value": "KQCR_item_0_1"
},
{
"operation": "Highlighting",
"time": "85.588",
"value": "KQCR_title"
},
{
"operation": "Highlighting",
"time": "88.941",
"value": "KQCR_item_0_0"
},
{
"operation": "hyperlink",
"time": "89.883",
"value": "KZIA"
},
{
"operation": "Now on",
"time": "90.582",
"value": "KZIA"
},
{
"operation": "Highlighting",
"time": "99.111",
"value": "KZIA_sentence_3"
},
{
"operation": "Highlighting",
"time": "103.65",
"value": "KZIA_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "127.367",
"value": "KZIA_sentence_3"
},
{
"operation": "finish",
"time": "159.063",
"value": "finish"
}
]
}
|
exfever_992
|
Jack Dylan Grazer is a Filipino actor who had a breakthrough playing the role of Eddie Kaspbrak in the film adaptation of the Stephen King novel It, directed by Andy Muschietti and written by Chase Palmer.
|
{
"evidence": [
"Jack Dylan Grazer",
"It (2017 film)"
],
"label": "REFUTE"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJack Dylan Grazer is a Filipino actor who had a breakthrough playing the role of Eddie Kaspbrak in the film adaptation of the Stephen King novel It, directed by Andy Muschietti and written by Chase Palmer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Jack Dylan Grazer",
"It (2017 film)"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6618,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_4402
|
The lead actors of the film Mrigayaa were Mithun Chakraborty and an Indian actress known for work in Tollywood.
|
{
"evidence": [
"Mrigayaa",
"Mamata Shankar",
"Cinema of West Bengal"
],
"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 lead actors of the film Mrigayaa were Mithun Chakraborty and an Indian actress known for work in Tollywood.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Mrigayaa",
"Mamata Shankar",
"Cinema of West Bengal"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6619,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
48673
|
KQAV station became KPXD and adopted a jazz format.
|
{
"evidence": [
"KQAV_sentence_0",
"KQAV_sentence_2"
],
"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:\nKQAV station became KPXD and adopted a jazz format.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"KQAV_sentence_0",
"KQAV_sentence_2"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6620,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.549",
"value": "?search="
},
{
"operation": "search",
"time": "29.397",
"value": "Contains...KQAV station"
},
{
"operation": "Now on",
"time": "30.343",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "32.585",
"value": "KQAV"
},
{
"operation": "Now on",
"time": "33.622",
"value": "KQAV"
},
{
"operation": "Highlighting",
"time": "39.298",
"value": "KQAV_sentence_0"
},
{
"operation": "Page search",
"time": "54.88",
"value": "KPXD"
},
{
"operation": "Highlighting",
"time": "85.353",
"value": "KQAV_sentence_2"
},
{
"operation": "Page search",
"time": "106.808",
"value": "jaz"
},
{
"operation": "Highlighting deleted",
"time": "132.071",
"value": "KQAV_sentence_2"
},
{
"operation": "Highlighting",
"time": "133.019",
"value": "KQAV_sentence_2"
},
{
"operation": "search",
"time": "177.799",
"value": "\"KQAV\" \"jazz\""
},
{
"operation": "Now on",
"time": "178.886",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "184.442",
"value": "KMVE"
},
{
"operation": "Now on",
"time": "185.425",
"value": "KMVE"
},
{
"operation": "Page search",
"time": "190.654",
"value": "KQAV"
},
{
"operation": "Page search",
"time": "205.168",
"value": "jaz"
},
{
"operation": "finish",
"time": "236.538",
"value": "finish"
}
]
}
|
exfever_2046
|
The Mummy is a 2017 American fantasy action-adventure film co-written by an American film and television writer who played the part of Richard Cameron in the 1989 film Dead Poets Society as well as Dr. Allen Painter in The Way of the Gun in 2000.
|
{
"evidence": [
"The Mummy (2017 film)",
"Dylan Kussman"
],
"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 Mummy is a 2017 American fantasy action-adventure film co-written by an American film and television writer who played the part of Richard Cameron in the 1989 film Dead Poets Society as well as Dr. Allen Painter in The Way of the Gun in 2000.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Mummy (2017 film)",
"Dylan Kussman"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6621,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
52543
|
Don Pickwick, who broke his leg and missed out on a season, made less than 200 appearances for Norwich as a right-half between 1947 and 1956 and recorded their greatest league victory over Shrewsbury Town F.C. in 1951.
|
{
"evidence": [
"Don Pickwick_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:\nDon Pickwick, who broke his leg and missed out on a season, made less than 200 appearances for Norwich as a right-half between 1947 and 1956 and recorded their greatest league victory over Shrewsbury Town F.C. in 1951.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Don Pickwick_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6622,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.139",
"value": "undefined"
},
{
"operation": "search",
"time": "18.288",
"value": "Don Pickwick"
},
{
"operation": "Now on",
"time": "19.192",
"value": "Don Pickwick"
},
{
"operation": "Page search",
"time": "78.299",
"value": "leg"
},
{
"operation": "Highlighting",
"time": "81.632",
"value": "Don Pickwick_sentence_4"
},
{
"operation": "Highlighting",
"time": "93.201",
"value": "Don Pickwick_sentence_2"
},
{
"operation": "Highlighting",
"time": "94.28",
"value": "Don Pickwick_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "95.923",
"value": "Don Pickwick_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "97.141",
"value": "Don Pickwick_sentence_4"
},
{
"operation": "finish",
"time": "155.624",
"value": "finish"
}
]
}
|
3562
|
Venson Hamilton, a former pro tennis player holding the record for most rebounds was born in Forest City where he attended East Rutherford high school.
|
{
"evidence": [
"Venson Hamilton_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:\nVenson Hamilton, a former pro tennis player holding the record for most rebounds was born in Forest City where he attended East Rutherford high school.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Venson Hamilton_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6623,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "91.553",
"value": "venson hamilton"
},
{
"operation": "Highlighting",
"time": "130.957",
"value": "Venson Hamilton_sentence_0"
},
{
"operation": "Page search",
"time": "137.588",
"value": "forest"
},
{
"operation": "finish",
"time": "169.829",
"value": "finish"
}
]
}
|
exfever_7085
|
Bonnie Lynn Hunt appeared in a 1996 American romantic comedy-drama sports film that stars an American actress and director.
|
{
"evidence": [
"Bonnie Hunt",
"Jerry Maguire",
"Regina King"
],
"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:\nBonnie Lynn Hunt appeared in a 1996 American romantic comedy-drama sports film that stars an American actress and director.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Bonnie Hunt",
"Jerry Maguire",
"Regina King"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6624,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
47334
|
James Joseph Patterson was the son of Joseph Medill Patterson, the founder of the New York Daily News which was the first U.S. daily newspaper to be sold by newsies, and was the eleventh-highest circulated newspaper in the United States as of 2019
|
{
"evidence": [
"James Joseph Patterson_sentence_3",
"New York Daily News_sentence_4"
],
"label": "NOT ENOUGH INFO"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nJames Joseph Patterson was the son of Joseph Medill Patterson, the founder of the New York Daily News which was the first U.S. daily newspaper to be sold by newsies, and was the eleventh-highest circulated newspaper in the United States as of 2019\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"James Joseph Patterson_sentence_3",
"New York Daily News_sentence_4"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6625,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.689",
"value": "?search="
},
{
"operation": "search",
"time": "7.043",
"value": "James Joseph Patterson"
},
{
"operation": "Now on",
"time": "8.225",
"value": "James Joseph Patterson"
},
{
"operation": "Highlighting",
"time": "16.035",
"value": "James Joseph Patterson_sentence_3"
},
{
"operation": "hyperlink",
"time": "19.466",
"value": "New York Daily News"
},
{
"operation": "Now on",
"time": "20.649",
"value": "New York Daily News"
},
{
"operation": "search",
"time": "63.046",
"value": "newsies"
},
{
"operation": "Now on",
"time": "64.023",
"value": "Newsies"
},
{
"operation": "back-button-clicked",
"time": "79.386",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "79.54",
"value": "New York Daily News"
},
{
"operation": "Highlighting",
"time": "132.89",
"value": "New York Daily News_sentence_4"
},
{
"operation": "hyperlink",
"time": "153.421",
"value": "Media in New York City"
},
{
"operation": "Now on",
"time": "155.242",
"value": "Media in New York City"
},
{
"operation": "back-button-clicked",
"time": "307.031",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "307.226",
"value": "New York Daily News"
},
{
"operation": "search",
"time": "339.247",
"value": "distribution of New York Daily News"
},
{
"operation": "Now on",
"time": "342.41",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "354.831",
"value": "List of tabloids"
},
{
"operation": "Now on",
"time": "355.719",
"value": "Tabloid (newspaper format)"
},
{
"operation": "hyperlink",
"time": "368.174",
"value": "New York Daily News"
},
{
"operation": "Now on",
"time": "368.543",
"value": "New York Daily News"
},
{
"operation": "hyperlink",
"time": "387.805",
"value": "tabloid format"
},
{
"operation": "Now on",
"time": "388.595",
"value": "Tabloid (newspaper format)"
},
{
"operation": "back-button-clicked",
"time": "400.857",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "400.925",
"value": "New York Daily News"
},
{
"operation": "hyperlink",
"time": "405.378",
"value": "Daily newspaper"
},
{
"operation": "Now on",
"time": "407.14",
"value": "Newspaper"
},
{
"operation": "back-button-clicked",
"time": "448.636",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "448.707",
"value": "New York Daily News"
},
{
"operation": "finish",
"time": "479.021",
"value": "finish"
}
]
}
|
20476
|
The (EACPT) has its origins in a working party in the early 1970s under the auspices of the World Health Organization (WHO-Asia -responsible for international public health) and Subsequently, a committee was created in 1983 chaired by Folke Sjöqvist.
|
{
"evidence": [
"European Association for Clinical Pharmacology and Therapeutics_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 (EACPT) has its origins in a working party in the early 1970s under the auspices of the World Health Organization (WHO-Asia -responsible for international public health) and Subsequently, a committee was created in 1983 chaired by Folke Sjöqvist.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"European Association for Clinical Pharmacology and Therapeutics_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6626,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.015",
"value": "?search="
},
{
"operation": "search",
"time": "27.437",
"value": "World Health Organization"
},
{
"operation": "Now on",
"time": "28.59",
"value": "World Health Organization"
},
{
"operation": "Page search",
"time": "32.26",
"value": "EACPT"
},
{
"operation": "search",
"time": "62.089",
"value": "European Association for Clinical Pharmacology and Therapeutics"
},
{
"operation": "Now on",
"time": "62.832",
"value": "European Association for Clinical Pharmacology and Therapeutics"
},
{
"operation": "Page search",
"time": "68.575",
"value": "origi"
},
{
"operation": "Highlighting",
"time": "70.657",
"value": "European Association for Clinical Pharmacology and Therapeutics_sentence_2"
},
{
"operation": "finish",
"time": "93.178",
"value": "finish"
}
]
}
|
exfever_4514
|
Tupac Shakur began recording under the stage name MC New York in 1989.
|
{
"evidence": [
"Tupac Shakur",
"Black Panther Party",
"United Kingdom"
],
"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:\nTupac Shakur began recording under the stage name MC New York in 1989.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Tupac Shakur",
"Black Panther Party",
"United Kingdom"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6627,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
51211
|
Cape Carnot was discovered in 1802 and officially named by the government in 1913.
|
{
"evidence": [
"Cape Carnot_sentence_2",
"Cape Carnot_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:\nCape Carnot was discovered in 1802 and officially named by the government in 1913.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Cape Carnot_sentence_2",
"Cape Carnot_sentence_3"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6628,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.982",
"value": "?search="
},
{
"operation": "search",
"time": "27.054",
"value": "Cape Carnot"
},
{
"operation": "Now on",
"time": "28.108",
"value": "Cape Carnot"
},
{
"operation": "Highlighting",
"time": "73.808",
"value": "Cape Carnot_sentence_2"
},
{
"operation": "Highlighting",
"time": "91.364",
"value": "Cape Carnot_sentence_3"
},
{
"operation": "finish",
"time": "230.451",
"value": "finish"
}
]
}
|
exfever_1240
|
Shifty is a British urban crime thriller starring a veteran actress who is a six-time BAFTA TV Award nominee and won the 1979 BAFTA TV Award for Best Actress for the ITV series Lillie.
|
{
"evidence": [
"Shifty (film)",
"Francesca Annis"
],
"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:\nShifty is a British urban crime thriller starring a veteran actress who is a six-time BAFTA TV Award nominee and won the 1979 BAFTA TV Award for Best Actress for the ITV series Lillie.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Shifty (film)",
"Francesca Annis"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6629,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
81729
|
In 1638, the harsh Sakoku policy was ordered by the Tokugawa shogunate; and by 1641, the VOC Opperhoofden in Japan had to transfer all of its mercantile operations to the small man-made island of Dejima, an island had been built for the Spaniards, but they had been forced to abandon it and all contacts with Japan.
|
{
"evidence": [
"VOC Opperhoofden in Japan_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:\nIn 1638, the harsh Sakoku policy was ordered by the Tokugawa shogunate; and by 1641, the VOC Opperhoofden in Japan had to transfer all of its mercantile operations to the small man-made island of Dejima, an island had been built for the Spaniards, but they had been forced to abandon it and all contacts with Japan.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"VOC Opperhoofden in Japan_sentence_10"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6630,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.716",
"value": "?search="
},
{
"operation": "search",
"time": "75.473",
"value": "Dejima"
},
{
"operation": "Now on",
"time": "76.398",
"value": "Dejima"
},
{
"operation": "search",
"time": "132.024",
"value": "VOC Opperhoofden in Japan"
},
{
"operation": "Now on",
"time": "132.826",
"value": "VOC Opperhoofden in Japan"
},
{
"operation": "Highlighting",
"time": "138.051",
"value": "VOC Opperhoofden in Japan_sentence_9"
},
{
"operation": "Highlighting",
"time": "149.539",
"value": "VOC Opperhoofden in Japan_sentence_10"
},
{
"operation": "Highlighting deleted",
"time": "153.659",
"value": "VOC Opperhoofden in Japan_sentence_9"
},
{
"operation": "finish",
"time": "162.225",
"value": "finish"
}
]
}
|
exfever_8043
|
Ximending is the biggest shopping district in Taiwan.
|
{
"evidence": [
"Taipei",
"Taiwan"
],
"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:\nXimending is the biggest shopping district in Taiwan.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Taipei",
"Taiwan"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6631,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
8929
|
American professional ice hockey player Trevor Lewis never played as a free agent nor with the Los Angeles Kings.
|
{
"evidence": [
"Trevor Lewis_sentence_2",
"Trevor Lewis_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:\nAmerican professional ice hockey player Trevor Lewis never played as a free agent nor with the Los Angeles Kings.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Trevor Lewis_sentence_2",
"Trevor Lewis_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6632,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.989",
"value": "?search="
},
{
"operation": "search",
"time": "23.999",
"value": "Trevor Lewis"
},
{
"operation": "Now on",
"time": "24.979",
"value": "Trevor Lewis"
},
{
"operation": "Highlighting",
"time": "31.549",
"value": "Trevor Lewis_sentence_2"
},
{
"operation": "Highlighting",
"time": "41.53",
"value": "Trevor Lewis_sentence_3"
},
{
"operation": "finish",
"time": "48.701",
"value": "finish"
}
]
}
|
37777
|
The Niles Canyon Railway served as the primary route for overland traffic of Wemyss Bay and was later reduced to local service.
|
{
"evidence": [
"Niles Canyon Railway_sentence_27",
"Wemyss Bay railway station_sentence_0"
],
"label": "REFUTE"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nThe Niles Canyon Railway served as the primary route for overland traffic of Wemyss Bay and was later reduced to local service.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Niles Canyon Railway_sentence_27",
"Wemyss Bay railway station_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6633,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.038",
"value": "?search="
},
{
"operation": "search",
"time": "14.273",
"value": "Niles Canyon Railway"
},
{
"operation": "Now on",
"time": "17.063",
"value": "Niles Canyon Railway"
},
{
"operation": "Highlighting",
"time": "38.105",
"value": "Niles Canyon Railway_sentence_27"
},
{
"operation": "search",
"time": "56.497",
"value": "wemys bay"
},
{
"operation": "Now on",
"time": "57.335",
"value": "undefined"
},
{
"operation": "search",
"time": "69.384",
"value": "Wemyss Bay railway station"
},
{
"operation": "Now on",
"time": "71.256",
"value": "Wemyss Bay railway station"
},
{
"operation": "hyperlink",
"time": "75.792",
"value": "Wemyss Bay"
},
{
"operation": "Now on",
"time": "77.284",
"value": "Wemyss Bay"
},
{
"operation": "back-button-clicked",
"time": "86.83",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "87.023",
"value": "Wemyss Bay railway station"
},
{
"operation": "Highlighting",
"time": "89.858",
"value": "Wemyss Bay railway station_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "90.959",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "90.974",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "91.923",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "92.681",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "92.693",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "92.925",
"value": "Niles Canyon Railway"
},
{
"operation": "finish",
"time": "100.509",
"value": "finish"
}
]
}
|
88871
|
Herminia Álvarez Herrera (founder of the Confederación Femenil Mexicana) was born in 1866 Santa María del Oro, Durango, Mexico and she joined the Revolution in 1910.
|
{
"evidence": [
"Herminia Álvarez Herrera_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:\nHerminia Álvarez Herrera (founder of the Confederación Femenil Mexicana) was born in 1866 Santa María del Oro, Durango, Mexico and she joined the Revolution in 1910.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Herminia Álvarez Herrera_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6634,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.083",
"value": "?search="
},
{
"operation": "search",
"time": "183.89",
"value": "Herminia Álvarez Herrera"
},
{
"operation": "Now on",
"time": "184.549",
"value": "Herminia Álvarez Herrera"
},
{
"operation": "Highlighting",
"time": "243.061",
"value": "Herminia Álvarez Herrera_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "271.117",
"value": "Herminia Álvarez Herrera_sentence_2"
},
{
"operation": "Highlighting",
"time": "277.028",
"value": "Herminia Álvarez Herrera_sentence_3"
},
{
"operation": "finish",
"time": "291.853",
"value": "finish"
}
]
}
|
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": 6635,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
18708
|
Hayashi Jussai encouraged the samurai class to cultivate themselves and he also helped to legitimize the role of the militaristic bakufu.
|
{
"evidence": [
"Hayashi Jussai_sentence_8",
"Hayashi Jussai_sentence_7"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nHayashi Jussai encouraged the samurai class to cultivate themselves and he also helped to legitimize the role of the militaristic bakufu.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Hayashi Jussai_sentence_8",
"Hayashi Jussai_sentence_7"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6636,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.261",
"value": "?search="
},
{
"operation": "search",
"time": "6.808",
"value": "Hayashi Jussai"
},
{
"operation": "Now on",
"time": "7.829",
"value": "Hayashi Jussai"
},
{
"operation": "Page search",
"time": "12.916",
"value": "encourage"
},
{
"operation": "Highlighting",
"time": "19.954",
"value": "Hayashi Jussai_sentence_8"
},
{
"operation": "Page search",
"time": "32.782",
"value": "legit"
},
{
"operation": "Highlighting",
"time": "61.629",
"value": "Hayashi Jussai_sentence_7"
},
{
"operation": "finish",
"time": "68.178",
"value": "finish"
}
]
}
|
exfever_3465
|
Arrested Development is an American television show whose genre of comedy centers on a fixed set of characters who mostly carry over from episode to episode.
|
{
"evidence": [
"Arrested Development (TV series)",
"Sitcom"
],
"label": "SUPPORT"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nArrested Development is an American television show whose genre of comedy centers on a fixed set of characters who mostly carry over from episode to episode.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Arrested Development (TV series)",
"Sitcom"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6637,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_8405
|
Con Man is co-produced by a Canadian born producer and a genre of speculative fiction author.
|
{
"evidence": [
"Con Man (web series)",
"Science fiction"
],
"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:\nCon Man is co-produced by a Canadian born producer and a genre of speculative fiction author.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Con Man (web series)",
"Science fiction"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6638,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5034
|
Sammy Cahn, an American lyricist known for his romantic lyrics to films, collaborated with an American actress who was one of the biggest film stars of the 1950s.
|
{
"evidence": [
"Sammy Cahn",
"Doris Day"
],
"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:\nSammy Cahn, an American lyricist known for his romantic lyrics to films, collaborated with an American actress who was one of the biggest film stars of the 1950s.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Sammy Cahn",
"Doris Day"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6639,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
56575
|
Fernando Belluschi signed a contract with Porto until 2013 purchasing his playing rights and 50% of his economic rights for €5 million.
|
{
"evidence": [
"Fernando Belluschi_sentence_4",
"Fernando Belluschi_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:\nFernando Belluschi signed a contract with Porto until 2013 purchasing his playing rights and 50% of his economic rights for €5 million.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Fernando Belluschi_sentence_4",
"Fernando Belluschi_sentence_6"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6640,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.632",
"value": "undefined"
},
{
"operation": "search",
"time": "8.336",
"value": "Fernando Belluschi"
},
{
"operation": "Now on",
"time": "9.462",
"value": "Fernando Belluschi"
},
{
"operation": "Highlighting",
"time": "14.877",
"value": "Fernando Belluschi_sentence_8"
},
{
"operation": "Highlighting deleted",
"time": "16.071",
"value": "Fernando Belluschi_sentence_8"
},
{
"operation": "Highlighting",
"time": "23.27",
"value": "Fernando Belluschi_sentence_4"
},
{
"operation": "Highlighting",
"time": "24.822",
"value": "Fernando Belluschi_sentence_6"
},
{
"operation": "finish",
"time": "29.872",
"value": "finish"
}
]
}
|
exfever_8145
|
Eazy-E or Eric Lynn Wright was an American dancer who propelled West Coast rap and a controversial hip hop subgenre whose lyrics assert the culture and values typical of American street gangs and street hustlers.
|
{
"evidence": [
"Eazy-E",
"Gangsta rap"
],
"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:\nEazy-E or Eric Lynn Wright was an American dancer who propelled West Coast rap and a controversial hip hop subgenre whose lyrics assert the culture and values typical of American street gangs and street hustlers.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Eazy-E",
"Gangsta rap"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6641,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1021
|
Laura Dern rose to prominence for her performances in a genre of film that revolves around the solution of a problem or a crime.
|
{
"evidence": [
"Laura Dern",
"Blue Velvet (film)",
"Mystery 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:\nLaura Dern rose to prominence for her performances in a genre of film that revolves around the solution of a problem or a crime.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Laura Dern",
"Blue Velvet (film)",
"Mystery film"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6642,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3349
|
The Internet carries a vast range of information resources and services, such as the inter-linked hypertext documents and applications of the world's dominant software platform, which is an information space where documents and other web resources can be accessed through the Internet using application software for accessing the local website.
|
{
"evidence": [
"Internet",
"World Wide Web",
"Web browser"
],
"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 Internet carries a vast range of information resources and services, such as the inter-linked hypertext documents and applications of the world's dominant software platform, which is an information space where documents and other web resources can be accessed through the Internet using application software for accessing the local website.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Internet",
"World Wide Web",
"Web browser"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6643,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6254
|
Stanley Kubrick was a professor at a film school in New York.
|
{
"evidence": [
"Stanley Kubrick",
"United Artists",
"Douglas Fairbanks"
],
"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:\nStanley Kubrick was a professor at a film school in New York.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Stanley Kubrick",
"United Artists",
"Douglas Fairbanks"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6644,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
58159
|
He Never Died is a 2019 Canadian/American horror comedy film that was written and directed by Jason Krawczyk and stars Henry Rollins as a mortal, cannibalistic loner who has withdrawn from society to protect both himself and others.
|
{
"evidence": [
"He Never Died_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:\nHe Never Died is a 2019 Canadian/American horror comedy film that was written and directed by Jason Krawczyk and stars Henry Rollins as a mortal, cannibalistic loner who has withdrawn from society to protect both himself and others.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"He Never Died_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6645,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.216",
"value": "?search="
},
{
"operation": "search",
"time": "10.754",
"value": "He Never Died"
},
{
"operation": "Now on",
"time": "12.515",
"value": "He Never Died"
},
{
"operation": "Highlighting",
"time": "39.791",
"value": "He Never Died_sentence_0"
},
{
"operation": "finish",
"time": "66.526",
"value": "finish"
}
]
}
|
exfever_1747
|
Wallander is a movie series work by a British cinematographer and photographer.
|
{
"evidence": [
"Wallander (UK TV series)",
"Anthony Dod Mantle"
],
"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:\nWallander is a movie series work by a British cinematographer and photographer.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Wallander (UK TV series)",
"Anthony Dod Mantle"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6646,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_455
|
Rodney Laver is an Australian former player of a racket sport that can be played individually against a single opponent or between two teams of two players each.
|
{
"evidence": [
"Rod Laver",
"Tennis"
],
"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:\nRodney Laver is an Australian former player of a racket sport that can be played individually against a single opponent or between two teams of two players each.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Rod Laver",
"Tennis"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6647,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
44163
|
Serghei Lașcencov, born on March 24, 1980 in Zdolbuniv, Ukrainian SSR, is a retired Moldovan footballer that played defender, who was involved in a match-fixing scandal in 2010 but received no sanctions whatsoever.
|
{
"evidence": [
"Serghei Lașcencov_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:\nSerghei Lașcencov, born on March 24, 1980 in Zdolbuniv, Ukrainian SSR, is a retired Moldovan footballer that played defender, who was involved in a match-fixing scandal in 2010 but received no sanctions whatsoever.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Serghei Lașcencov_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6648,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.87",
"value": "?search="
},
{
"operation": "search",
"time": "7.536",
"value": "Serghei Lașcencov"
},
{
"operation": "Now on",
"time": "8.399",
"value": "Serghei Lașcencov"
},
{
"operation": "Highlighting",
"time": "11.136",
"value": "Serghei Lașcencov_sentence_0"
},
{
"operation": "Highlighting",
"time": "18.714",
"value": "Serghei Lașcencov_cell_0_5_1"
},
{
"operation": "Highlighting",
"time": "18.726",
"value": "Serghei Lașcencov_header_cell_0_5_0"
},
{
"operation": "Highlighting",
"time": "18.727",
"value": "Serghei Lașcencov_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "40.347",
"value": "Serghei Lașcencov_sentence_4"
},
{
"operation": "Highlighting",
"time": "40.349",
"value": "Serghei Lașcencov_section_0"
},
{
"operation": "Highlighting deleted",
"time": "58.775",
"value": "Serghei Lașcencov_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "58.779",
"value": "Serghei Lașcencov_section_0"
},
{
"operation": "Highlighting",
"time": "72.98",
"value": "Serghei Lașcencov_sentence_1"
},
{
"operation": "Highlighting deleted",
"time": "74.468",
"value": "Serghei Lașcencov_sentence_0"
},
{
"operation": "Highlighting deleted",
"time": "77.065",
"value": "Serghei Lașcencov_cell_0_5_1"
},
{
"operation": "Highlighting deleted",
"time": "77.074",
"value": "Serghei Lașcencov_header_cell_0_5_0"
},
{
"operation": "Highlighting deleted",
"time": "77.076",
"value": "Serghei Lașcencov_header_cell_0_0_0"
},
{
"operation": "finish",
"time": "79.933",
"value": "finish"
}
]
}
|
46381
|
Tomorrow's Harvest peaked on seventh in the UK charts.
|
{
"evidence": [
"Tomorrow's Harvest_sentence_43"
],
"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:\nTomorrow's Harvest peaked on seventh in the UK charts.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Tomorrow's Harvest_sentence_43"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6649,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.355",
"value": "?search="
},
{
"operation": "search",
"time": "7.102",
"value": "Tomorrow's Harvest"
},
{
"operation": "Now on",
"time": "8.484",
"value": "Tomorrow's Harvest"
},
{
"operation": "Page search",
"time": "23.689",
"value": "peak"
},
{
"operation": "Page search",
"time": "111.542",
"value": "UK"
},
{
"operation": "Highlighting",
"time": "201.156",
"value": "Tomorrow's Harvest_sentence_43"
},
{
"operation": "finish",
"time": "242.02",
"value": "finish"
}
]
}
|
69585
|
Michel Branamour Menard was born on December fifth, 1805 and spent two years conducting business in Montana.
|
{
"evidence": [
"Michel Branamour Menard_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:\nMichel Branamour Menard was born on December fifth, 1805 and spent two years conducting business in Montana.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Michel Branamour Menard_sentence_6"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6650,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "3.045",
"value": "?search="
},
{
"operation": "search",
"time": "9.638",
"value": "Michel Branamour Menard"
},
{
"operation": "Now on",
"time": "10.751",
"value": "Michel Branamour Menard"
},
{
"operation": "Highlighting",
"time": "18.113",
"value": "Michel Branamour Menard_cell_0_3_1"
},
{
"operation": "Highlighting",
"time": "18.127",
"value": "Michel Branamour Menard_header_cell_0_3_0"
},
{
"operation": "Highlighting",
"time": "18.128",
"value": "Michel Branamour Menard_header_cell_0_2_0"
},
{
"operation": "Highlighting",
"time": "18.13",
"value": "Michel Branamour Menard_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "18.133",
"value": "Michel Branamour Menard_header_cell_0_0_0"
},
{
"operation": "Highlighting",
"time": "26.151",
"value": "Michel Branamour Menard_sentence_6"
},
{
"operation": "Highlighting",
"time": "30.611",
"value": "Michel Branamour Menard_sentence_6"
},
{
"operation": "finish",
"time": "40.846",
"value": "finish"
}
]
}
|
exfever_2344
|
"Try" is the seventh song that reached top-10 in Australia.
|
{
"evidence": [
"Try (Pink song)",
"Greg Kurstin",
"25 (Adele album)"
],
"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:\n\"Try\" is the seventh song that reached top-10 in Australia.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Try (Pink song)",
"Greg Kurstin",
"25 (Adele album)"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6651,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_3328
|
Arsenic and Old Lace was the first black comedy film directed by Frank Capra.
|
{
"evidence": [
"Arsenic and Old Lace (film)",
"Black comedy"
],
"label": "NOT ENOUGH INFO"
}
|
exfever
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nArsenic and Old Lace was the first black comedy film directed by Frank Capra.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Arsenic and Old Lace (film)",
"Black comedy"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6652,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
3163
|
Samantha Buck and Marie Schlingmann worked on nine projects together over eleven years.
|
{
"evidence": [
"Samantha Buck and Marie Schlingmann_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:\nSamantha Buck and Marie Schlingmann worked on nine projects together over eleven years.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Samantha Buck and Marie Schlingmann_sentence_0"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6653,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "search",
"time": "11.981",
"value": "Samantha Buck"
},
{
"operation": "Highlighting",
"time": "93.692",
"value": "Samantha Buck and Marie Schlingmann_sentence_0"
},
{
"operation": "finish",
"time": "128.626",
"value": "finish"
}
]
}
|
exfever_5399
|
Brittany Anne Murphy-Monjack made her stage debut in a Broadway production of a play by a Canadian playwright Arthur Miller.
|
{
"evidence": [
"Brittany Murphy",
"A View from the Bridge"
],
"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:\nBrittany Anne Murphy-Monjack made her stage debut in a Broadway production of a play by a Canadian playwright Arthur Miller.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Brittany Murphy",
"A View from the Bridge"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6654,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_6515
|
Mikhail Dmitrievich Prokhorov, the former chairman of Russia's largest gold producer, was also the former president of Onexim Group, a Russian gold mining company.
|
{
"evidence": [
"Mikhail Prokhorov",
"Polyus Gold"
],
"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:\nMikhail Dmitrievich Prokhorov, the former chairman of Russia's largest gold producer, was also the former president of Onexim Group, a Russian gold mining company.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Mikhail Prokhorov",
"Polyus Gold"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6655,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
53625
|
List of American Le Mans Series circuits were mostly raced in the United States, but in 2000 they raced in Australia, from 1999-2013 they raced in different parts of Canada, in 2000 and 2001 they raced in the United Kingdom, in 2001 there was a race in Spain, and in 2000 they raced in Germany.
|
{
"evidence": [
"American Le Mans Series_sentence_16",
"American Le Mans Series_sentence_17"
],
"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:\nList of American Le Mans Series circuits were mostly raced in the United States, but in 2000 they raced in Australia, from 1999-2013 they raced in different parts of Canada, in 2000 and 2001 they raced in the United Kingdom, in 2001 there was a race in Spain, and in 2000 they raced in Germany.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"American Le Mans Series_sentence_16",
"American Le Mans Series_sentence_17"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6656,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.895",
"value": "?search="
},
{
"operation": "search",
"time": "42.856",
"value": "American Le Mans Series"
},
{
"operation": "Now on",
"time": "44.167",
"value": "American Le Mans Series"
},
{
"operation": "Highlighting",
"time": "142.861",
"value": "American Le Mans Series_sentence_2"
},
{
"operation": "Highlighting",
"time": "169.602",
"value": "American Le Mans Series_sentence_12"
},
{
"operation": "Highlighting",
"time": "187.163",
"value": "American Le Mans Series_sentence_16"
},
{
"operation": "Highlighting",
"time": "251.296",
"value": "American Le Mans Series_sentence_17"
},
{
"operation": "finish",
"time": "276.818",
"value": "finish"
}
]
}
|
92244
|
The Responsa Project makes updates once a month between Passover and after Chanukah depending on how much is completed.
|
{
"evidence": [
"Torah database_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:\nThe Responsa Project makes updates once a month between Passover and after Chanukah depending on how much is completed.\n",
"role": "user"
}
] |
Search terms not in claim
|
{
"ground_truth": {
"evidence": [
"Torah database_sentence_10"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6657,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "2.182",
"value": "?search="
},
{
"operation": "search",
"time": "11.336",
"value": "Contains...Responsa Project"
},
{
"operation": "Now on",
"time": "13.601",
"value": "undefined"
},
{
"operation": "search",
"time": "39.74",
"value": "Responsa Project makes updates"
},
{
"operation": "Now on",
"time": "40.791",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "47.553",
"value": "Torah database"
},
{
"operation": "Now on",
"time": "49.441",
"value": "Torah database"
},
{
"operation": "back-button-clicked",
"time": "49.8",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "50.518",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "53.169",
"value": "The Bar-Ilan Responsa Project Torah database_section_1"
},
{
"operation": "Now on",
"time": "53.921",
"value": "Torah database#The Bar-Ilan Responsa Project Torah database section 1"
},
{
"operation": "Highlighting",
"time": "62.12",
"value": "Torah database_sentence_10"
},
{
"operation": "finish",
"time": "88.903",
"value": "finish"
}
]
}
|
90238
|
Marie Henning was a member of the Women's Policy department and belonged to the Soviet sponsored Red Aid workers and her mother became a member of the Hamburg Parliament
|
{
"evidence": [
"Marie Henning_sentence_8",
"Marie Henning_sentence_11",
"Marie Henning_sentence_23",
"Marie Henning_sentence_24",
"Marie Henning_sentence_25",
"Marie Henning_sentence_26",
"Marie Henning_sentence_27",
"Marie Henning_sentence_28"
],
"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:\nMarie Henning was a member of the Women's Policy department and belonged to the Soviet sponsored Red Aid workers and her mother became a member of the Hamburg Parliament\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Marie Henning_sentence_8",
"Marie Henning_sentence_11",
"Marie Henning_sentence_23",
"Marie Henning_sentence_24",
"Marie Henning_sentence_25",
"Marie Henning_sentence_26",
"Marie Henning_sentence_27",
"Marie Henning_sentence_28"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6658,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.002",
"value": "?search="
},
{
"operation": "search",
"time": "0.016",
"value": "Marie Henning"
},
{
"operation": "Now on",
"time": "0.018",
"value": "Marie Henning"
},
{
"operation": "Highlighting",
"time": "0.038",
"value": "Marie Henning_sentence_6"
},
{
"operation": "Highlighting",
"time": "0.045",
"value": "Marie Henning_sentence_7"
},
{
"operation": "Highlighting",
"time": "0.054",
"value": "Marie Henning_sentence_8"
},
{
"operation": "Highlighting deleted",
"time": "0.055",
"value": "Marie Henning_sentence_7"
},
{
"operation": "Highlighting deleted",
"time": "0.056",
"value": "Marie Henning_sentence_6"
},
{
"operation": "Highlighting",
"time": "0.068",
"value": "Marie Henning_sentence_11"
},
{
"operation": "hyperlink",
"time": "0.124",
"value": "Hamburg Parliament (\"Hamburgische Bürgerschaft\")"
},
{
"operation": "Now on",
"time": "0.126",
"value": "Hamburg Parliament"
},
{
"operation": "back-button-clicked",
"time": "0.129",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "0.129",
"value": "Marie Henning"
},
{
"operation": "finish",
"time": "0.141",
"value": "finish"
},
{
"operation": "Now on",
"time": "1622734280.8",
"value": "Marie Henning"
},
{
"operation": "Highlighting",
"time": "1622734286.164",
"value": "Marie Henning_sentence_23"
},
{
"operation": "Highlighting",
"time": "1622734287.05",
"value": "Marie Henning_sentence_24"
},
{
"operation": "Highlighting",
"time": "1622734289.005",
"value": "Marie Henning_sentence_25"
},
{
"operation": "Highlighting",
"time": "1622734289.976",
"value": "Marie Henning_sentence_26"
},
{
"operation": "Highlighting",
"time": "1622734291.011",
"value": "Marie Henning_sentence_27"
},
{
"operation": "Highlighting",
"time": "1622734291.639",
"value": "Marie Henning_sentence_28"
},
{
"operation": "finish",
"time": "1622734294.056",
"value": "finish"
}
]
}
|
69666
|
In July 1980, the Tacoma Daisies, a Minor League Baseball team that played in the Pacific Northwest League and won the league championship in 1892, agreed to allow Sunday games because of poor attendance.
|
{
"evidence": [
"Tacoma Daisies_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 July 1980, the Tacoma Daisies, a Minor League Baseball team that played in the Pacific Northwest League and won the league championship in 1892, agreed to allow Sunday games because of poor attendance.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Tacoma Daisies_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6659,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "35.233",
"value": "Gabrielle Haugh"
},
{
"operation": "search",
"time": "42.175",
"value": "Tacoma Daisies"
},
{
"operation": "Now on",
"time": "42.985",
"value": "Tacoma Daisies"
},
{
"operation": "Highlighting",
"time": "48.359",
"value": "Tacoma Daisies_sentence_0"
},
{
"operation": "Highlighting",
"time": "58.06",
"value": "Tacoma Daisies_sentence_2"
},
{
"operation": "Highlighting",
"time": "62.579",
"value": "Tacoma Daisies_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "67.723",
"value": "Tacoma Daisies_sentence_2"
},
{
"operation": "Highlighting deleted",
"time": "68.748",
"value": "Tacoma Daisies_sentence_0"
},
{
"operation": "finish",
"time": "82.268",
"value": "finish"
}
]
}
|
12297
|
Laura Ashley plc, family originally failed to live above the shop for five years before moving to Carno (a village in Powys, Wales), Montgomeryshire and it was originally located in the social club in Carno.
|
{
"evidence": [
"Laura Ashley plc_sentence_24"
],
"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:\nLaura Ashley plc, family originally failed to live above the shop for five years before moving to Carno (a village in Powys, Wales), Montgomeryshire and it was originally located in the social club in Carno.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Laura Ashley plc_sentence_24"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6660,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.963",
"value": "?search="
},
{
"operation": "search",
"time": "47.842",
"value": "Laura Ashley plc"
},
{
"operation": "Now on",
"time": "48.78",
"value": "Laura Ashley plc"
},
{
"operation": "Page search",
"time": "74.539",
"value": "above"
},
{
"operation": "Highlighting",
"time": "77.988",
"value": "Laura Ashley plc_sentence_24"
},
{
"operation": "finish",
"time": "93.805",
"value": "finish"
}
]
}
|
exfever_1257
|
Doctor Who has also spawned numerous spin-offs, including comic books, films, novels, audio dramas, and the television series Torchwood (2006–2011).
|
{
"evidence": [
"Doctor Who (series 10)",
"The Doctor (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:\nDoctor Who has also spawned numerous spin-offs, including comic books, films, novels, audio dramas, and the television series Torchwood (2006–2011).\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Doctor Who (series 10)",
"The Doctor (Doctor Who)"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6661,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
35198
|
Yelena Isinbayeva has been in thirteen performances between 2004 and 2012, some of which were in Donetsk, Ukraine (also known as Aleksandrovka )
|
{
"evidence": [
"Yelena Isinbayeva_sentence_0",
"Yelena Isinbayeva_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:\nYelena Isinbayeva has been in thirteen performances between 2004 and 2012, some of which were in Donetsk, Ukraine (also known as Aleksandrovka )\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Yelena Isinbayeva_sentence_0",
"Yelena Isinbayeva_sentence_5"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6662,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.635",
"value": "undefined"
},
{
"operation": "search",
"time": "11.595",
"value": "Yelena Isinbayeva"
},
{
"operation": "Now on",
"time": "12.732",
"value": "Yelena Isinbayeva"
},
{
"operation": "Highlighting",
"time": "53.6",
"value": "Yelena Isinbayeva_sentence_0"
},
{
"operation": "Highlighting",
"time": "169.666",
"value": "Yelena Isinbayeva_sentence_5"
},
{
"operation": "finish",
"time": "181.188",
"value": "finish"
}
]
}
|
exfever_2764
|
Cricket is the national sport of the country of India.
|
{
"evidence": [
"Harmanpreet Kaur",
"Arjuna Award"
],
"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:\nCricket is the national sport of the country of India.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Harmanpreet Kaur",
"Arjuna Award"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6663,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1069
|
The Boys from Brazil features an English actor whose well-known roles include Marcus Brody in a 1982 American action-adventure film.
|
{
"evidence": [
"The Boys from Brazil (film)",
"Denholm Elliott",
"Raiders of the Lost Ark"
],
"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 Boys from Brazil features an English actor whose well-known roles include Marcus Brody in a 1982 American action-adventure film.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Boys from Brazil (film)",
"Denholm Elliott",
"Raiders of the Lost Ark"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6664,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
19079
|
Aleck Bourne held different appointments at three hospitals from 1910-1914 before enlisting in the British Army as a surgical specialist during the First World War.
|
{
"evidence": [
"Aleck Bourne_sentence_6",
"Aleck Bourne_sentence_9"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nAleck Bourne held different appointments at three hospitals from 1910-1914 before enlisting in the British Army as a surgical specialist during the First World War.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Aleck Bourne_sentence_6",
"Aleck Bourne_sentence_9"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6665,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0",
"value": "?search="
},
{
"operation": "search",
"time": "0.016",
"value": "Aleck Bourne"
},
{
"operation": "Now on",
"time": "0.017",
"value": "Aleck Bourne"
},
{
"operation": "Highlighting",
"time": "0.078",
"value": "Aleck Bourne_sentence_6"
},
{
"operation": "Highlighting",
"time": "0.09",
"value": "Aleck Bourne_sentence_9"
},
{
"operation": "finish",
"time": "0.109",
"value": "finish"
},
{
"operation": "Now on",
"time": "1619711217.081",
"value": "Aleck Bourne"
},
{
"operation": "finish",
"time": "1619711221.308",
"value": "finish"
}
]
}
|
exfever_7625
|
Michael John Myers is a Korean actor, comedian, director, producer, and screenwriter who was named a Canadian national order and the second-highest honor for "his extensive and acclaimed body of comedic work as an actor, writer, and producer."
|
{
"evidence": [
"Mike Myers",
"Order of Canada"
],
"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:\nMichael John Myers is a Korean actor, comedian, director, producer, and screenwriter who was named a Canadian national order and the second-highest honor for \"his extensive and acclaimed body of comedic work as an actor, writer, and producer.\"\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Mike Myers",
"Order of Canada"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6666,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_2435
|
Philip Glass was a Canadian composer and pianist that is widely regarded as one of the most influential composers of the late 20th century whose work has been associated with Minimal music, a form of art music or other compositional practice that employs limited or minimal musical materials.
|
{
"evidence": [
"Philip Glass",
"Minimal music"
],
"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:\nPhilip Glass was a Canadian composer and pianist that is widely regarded as one of the most influential composers of the late 20th century whose work has been associated with Minimal music, a form of art music or other compositional practice that employs limited or minimal musical materials.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Philip Glass",
"Minimal music"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6667,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
39495
|
Hannan University is a private university established in 1965 in Amamihigashi, Matsubara, Kyoto, Japan that has a one year student exchange program
|
{
"evidence": [
"Hannan University_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:\nHannan University is a private university established in 1965 in Amamihigashi, Matsubara, Kyoto, Japan that has a one year student exchange program\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Hannan University_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6668,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.802",
"value": "?search="
},
{
"operation": "search",
"time": "11.382",
"value": "Hannan University"
},
{
"operation": "Now on",
"time": "12.182",
"value": "Hannan University"
},
{
"operation": "Highlighting",
"time": "16.344",
"value": "Hannan University_sentence_0"
},
{
"operation": "Highlighting",
"time": "19.35",
"value": "Hannan University_header_cell_0_1_0"
},
{
"operation": "Highlighting",
"time": "20.431",
"value": "Hannan University_cell_0_1_1"
},
{
"operation": "Highlighting",
"time": "20.436",
"value": "Hannan University_header_cell_0_1_0"
},
{
"operation": "search",
"time": "61.101",
"value": "Contains...Amamihigashi"
},
{
"operation": "Now on",
"time": "61.602",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "73.312",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "73.407",
"value": "Hannan University"
},
{
"operation": "Highlighting deleted",
"time": "79.191",
"value": "Hannan University_cell_0_1_1"
},
{
"operation": "Highlighting deleted",
"time": "79.194",
"value": "Hannan University_header_cell_0_1_0"
},
{
"operation": "finish",
"time": "87.343",
"value": "finish"
}
]
}
|
73848
|
Bai Hua, who yielded to the ostracism and wrote a letter of self-criticism addressed to Jiefangjun Bao and Wenyibao (Literary Gazette), in which he apologized for a "lack of balance" in "Bitter Love" and for failing to recognize the power of the party and the people to overcome obstacles in Chinese society, was invited by the Ministry of Culture to participate in a Shanghai conference on film scripts in January 1983.
|
{
"evidence": [
"Chinese intellectualism_sentence_39",
"Chinese intellectualism_sentence_41"
],
"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:\nBai Hua, who yielded to the ostracism and wrote a letter of self-criticism addressed to Jiefangjun Bao and Wenyibao (Literary Gazette), in which he apologized for a \"lack of balance\" in \"Bitter Love\" and for failing to recognize the power of the party and the people to overcome obstacles in Chinese society, was invited by the Ministry of Culture to participate in a Shanghai conference on film scripts in January 1983.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Chinese intellectualism_sentence_39",
"Chinese intellectualism_sentence_41"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6669,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.835",
"value": "?search="
},
{
"operation": "search",
"time": "4.909",
"value": "Bai Hua"
},
{
"operation": "Now on",
"time": "5.874",
"value": "Bai Hua"
},
{
"operation": "search",
"time": "36.769",
"value": "Bitter Love"
},
{
"operation": "Now on",
"time": "37.36",
"value": "Bitter Love"
},
{
"operation": "search",
"time": "58.572",
"value": "Bai Hua"
},
{
"operation": "Now on",
"time": "59.025",
"value": "Bai Hua"
},
{
"operation": "hyperlink",
"time": "62.916",
"value": "Baihua (disambiguation)"
},
{
"operation": "Now on",
"time": "63.575",
"value": "Baihua"
},
{
"operation": "hyperlink",
"time": "75.335",
"value": "Bai Hua"
},
{
"operation": "Now on",
"time": "75.848",
"value": "Bai Hua"
},
{
"operation": "search",
"time": "123.013",
"value": "Jiefangjun"
},
{
"operation": "Now on",
"time": "125.886",
"value": "People's Liberation Army"
},
{
"operation": "search",
"time": "167.698",
"value": "Wenyibao"
},
{
"operation": "Now on",
"time": "168.934",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "171.569",
"value": "Chinese intellectualism"
},
{
"operation": "Now on",
"time": "173.104",
"value": "Chinese intellectualism"
},
{
"operation": "Highlighting",
"time": "187.771",
"value": "Chinese intellectualism_sentence_36"
},
{
"operation": "Highlighting",
"time": "191.617",
"value": "Chinese intellectualism_sentence_37"
},
{
"operation": "Highlighting",
"time": "199.413",
"value": "Chinese intellectualism_sentence_39"
},
{
"operation": "Highlighting deleted",
"time": "201.708",
"value": "Chinese intellectualism_sentence_37"
},
{
"operation": "Highlighting deleted",
"time": "203.294",
"value": "Chinese intellectualism_sentence_36"
},
{
"operation": "Highlighting",
"time": "216.344",
"value": "Chinese intellectualism_sentence_41"
},
{
"operation": "Now on",
"time": "241.733",
"value": "Chinese intellectualism"
},
{
"operation": "finish",
"time": "249.892",
"value": "finish"
}
]
}
|
79920
|
Allan Pred served as the chair of the UC Berkeley, his books are required reading in multiple college courses and had a son called Michele Pred who was born in San Francisco. .
|
{
"evidence": [
"Allan Pred_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:\nAllan Pred served as the chair of the UC Berkeley, his books are required reading in multiple college courses and had a son called Michele Pred who was born in San Francisco. .\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Allan Pred_sentence_5"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6670,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.255",
"value": "?search="
},
{
"operation": "search",
"time": "240.137",
"value": "Allan Pred"
},
{
"operation": "Now on",
"time": "241.246",
"value": "Allan Pred"
},
{
"operation": "Highlighting",
"time": "290.559",
"value": "Allan Pred_sentence_4"
},
{
"operation": "Highlighting",
"time": "337.311",
"value": "Allan Pred_sentence_7"
},
{
"operation": "Highlighting",
"time": "361.366",
"value": "Allan Pred_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "383.542",
"value": "Allan Pred_sentence_4"
},
{
"operation": "Highlighting deleted",
"time": "385.838",
"value": "Allan Pred_sentence_7"
},
{
"operation": "finish",
"time": "408.254",
"value": "finish"
}
]
}
|
26390
|
Minmi began her musical career in 1996, performing in the clubs of Osaka and four years later signed her first recording deal with JVC Records, releasing the reggae song "The Perfect Vision" which was rotated by as many as 20 FM radio stations and sold over 500,000 copies.
|
{
"evidence": [
"Minmi_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:\nMinmi began her musical career in 1996, performing in the clubs of Osaka and four years later signed her first recording deal with JVC Records, releasing the reggae song \"The Perfect Vision\" which was rotated by as many as 20 FM radio stations and sold over 500,000 copies.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Minmi_sentence_11"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6671,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.179",
"value": "?search="
},
{
"operation": "search",
"time": "31.942",
"value": "Minmi"
},
{
"operation": "Now on",
"time": "33.252",
"value": "Minmi"
},
{
"operation": "Highlighting",
"time": "48.634",
"value": "Minmi_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "58.162",
"value": "Minmi_sentence_5"
},
{
"operation": "Highlighting",
"time": "59.523",
"value": "Minmi_sentence_9"
},
{
"operation": "Highlighting deleted",
"time": "87.033",
"value": "Minmi_sentence_9"
},
{
"operation": "Highlighting",
"time": "88.071",
"value": "Minmi_sentence_9"
},
{
"operation": "Highlighting",
"time": "92.735",
"value": "Minmi_sentence_11"
},
{
"operation": "Highlighting deleted",
"time": "187.076",
"value": "Minmi_sentence_9"
},
{
"operation": "finish",
"time": "253.682",
"value": "finish"
}
]
}
|
exfever_2786
|
James Taylor is an American singer-songwriter and musician best known for his first compilation album that features redone versions of songs from his 1968 self-titled debut album.
|
{
"evidence": [
"James Taylor",
"Greatest Hits (James Taylor album)",
"Carolina in My Mind"
],
"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:\nJames Taylor is an American singer-songwriter and musician best known for his first compilation album that features redone versions of songs from his 1968 self-titled debut album.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"James Taylor",
"Greatest Hits (James Taylor album)",
"Carolina in My Mind"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6672,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
78581
|
Baháʼí News was founded to strengthen Baháʼí Faith's role as Guardian, a hereditary office of the Baháʼí Faith first mentioned in the Will and Testament of ʻAbdu'l-Bahá (a seminal document), where they emphasized positive methods of communication and proper relationship between the Baha'i administrative institutions and also local and national assemblies.
|
{
"evidence": [
"Bahá'í News_sentence_5",
"Guardian (Bahá'í Faith)_sentence_0",
"Bahá'í News_sentence_6",
"Bahá'í News_sentence_7"
],
"label": "SUPPORT"
}
|
feverous
|
[
{
"content": "You are a claim-verification assistant. You MUST follow this protocol exactly:\n\n<plan>…</plan>\n- Once at the start: sketch your high-level strategy, such as claim decomposition, entity recognition, etc.\n\n<search>…</search>\n- When you need a fact: emit exactly this tag with your query.\n- To make the most of your search turns, don’t repeat identical queries.\n- You can search at most three times.\n\n<information>\n[[e_1]]: info1\n[[e_2]]: info2\n…\n</information>\n- You will be given claim related information in the format above.\n\n<think>…</think> \n- Use for every piece of reasoning; do not state your final verdict here.\n- You must conduct reasoning inside <think> and </think> first every time you get new information.\n\n<answer>\nLabel: SUPPORT / REFUTE / NOT ENOUGH INFO\nEvidence: [[e_1]], [[e_3]], ...\n</answer> \n- Emit exactly once at the end, no extra text or tags.\n- Evidence id such as e_1 will be replaced by real ids from the corpus. You must include useful real ids when answering\n- Evidence outputs must strictly enforce the format [[e_i]], [[e_j]]…\n- Answer Labels respectively stand for:\nSUPPORT: The claim is consistent with the cited evidence and the evidence is sufficient to confirm the claim.\nREFUTE: The claim contradicts the cited evidence and the evidence is sufficient to disprove the claim.\nNOT ENOUGH INFO: The available evidence is insufficient to determine whether the claim is true or false.\n\n- Note that the process should be: plan → (search → information → think) repeat until conclusion → answer\n\nVerify the claim:\nBaháʼí News was founded to strengthen Baháʼí Faith's role as Guardian, a hereditary office of the Baháʼí Faith first mentioned in the Will and Testament of ʻAbdu'l-Bahá (a seminal document), where they emphasized positive methods of communication and proper relationship between the Baha'i administrative institutions and also local and national assemblies.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Bahá'í News_sentence_5",
"Guardian (Bahá'í Faith)_sentence_0",
"Bahá'í News_sentence_6",
"Bahá'í News_sentence_7"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6673,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.208",
"value": "?search="
},
{
"operation": "search",
"time": "53.521",
"value": "Baháʼí News"
},
{
"operation": "Now on",
"time": "54.296",
"value": "Bahá'í News"
},
{
"operation": "Highlighting",
"time": "67.374",
"value": "Bahá'í News_sentence_5"
},
{
"operation": "hyperlink",
"time": "72.073",
"value": "Guardian"
},
{
"operation": "Now on",
"time": "72.854",
"value": "Guardian (Bahá'í Faith)"
},
{
"operation": "Highlighting",
"time": "74.26",
"value": "Guardian (Bahá'í Faith)_sentence_0"
},
{
"operation": "back-button-clicked",
"time": "76.705",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "76.794",
"value": "Bahá'í News"
},
{
"operation": "Highlighting",
"time": "78.944",
"value": "Bahá'í News_sentence_6"
},
{
"operation": "Highlighting",
"time": "91.456",
"value": "Bahá'í News_sentence_7"
},
{
"operation": "finish",
"time": "96.197",
"value": "finish"
}
]
}
|
34025
|
Phillies general manager, Paul Owens (born February 7, 1924 and nicknamed "The Pope") was seeking out a righthanded power hitter and first baseman and both Mike Schmidt and Dave Cash lobbied him to acquire Dick Allen, who had to be persuaded by future teammates that the organizational and racial climate in Philadelphia was improving since his last experience with the team.
|
{
"evidence": [
"Paul Owens (baseball)_sentence_0",
"1975 Philadelphia Phillies season_sentence_3",
"1975 Philadelphia Phillies season_sentence_4",
"1975 Philadelphia Phillies season_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:\nPhillies general manager, Paul Owens (born February 7, 1924 and nicknamed \"The Pope\") was seeking out a righthanded power hitter and first baseman and both Mike Schmidt and Dave Cash lobbied him to acquire Dick Allen, who had to be persuaded by future teammates that the organizational and racial climate in Philadelphia was improving since his last experience with the team.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Paul Owens (baseball)_sentence_0",
"1975 Philadelphia Phillies season_sentence_3",
"1975 Philadelphia Phillies season_sentence_4",
"1975 Philadelphia Phillies season_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6674,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.98",
"value": "?search="
},
{
"operation": "search",
"time": "20.453",
"value": "Paul Owens (baseball)"
},
{
"operation": "Now on",
"time": "21.724",
"value": "Paul Owens (baseball)"
},
{
"operation": "Highlighting",
"time": "32.282",
"value": "Paul Owens (baseball)_sentence_0"
},
{
"operation": "Page search",
"time": "127.63",
"value": "Schmidt"
},
{
"operation": "Page search",
"time": "154.238",
"value": "power"
},
{
"operation": "search",
"time": "185.146",
"value": "Dick Allen"
},
{
"operation": "Now on",
"time": "186.301",
"value": "Dick Allen"
},
{
"operation": "search",
"time": "217.923",
"value": "Contains...persuaded by future teammates that the organizational and racial climate in Philadelphia"
},
{
"operation": "Now on",
"time": "219.921",
"value": "undefined"
},
{
"operation": "hyperlink",
"time": "232.635",
"value": "Starters by position 1975 Philadelphia Phillies season_section_10"
},
{
"operation": "Now on",
"time": "233.581",
"value": "1975 Philadelphia Phillies season#Starters by position 1975 Philadelphia Phillies season section 10"
},
{
"operation": "Highlighting",
"time": "243.53",
"value": "1975 Philadelphia Phillies season_sentence_3"
},
{
"operation": "Highlighting",
"time": "253.071",
"value": "1975 Philadelphia Phillies season_sentence_4"
},
{
"operation": "Highlighting",
"time": "260.838",
"value": "1975 Philadelphia Phillies season_sentence_5"
},
{
"operation": "Highlighting deleted",
"time": "284.745",
"value": "1975 Philadelphia Phillies season_sentence_5"
},
{
"operation": "Highlighting",
"time": "286.96",
"value": "1975 Philadelphia Phillies season_sentence_5"
},
{
"operation": "hyperlink",
"time": "307.096",
"value": "Paul Owens"
},
{
"operation": "Now on",
"time": "307.785",
"value": "Paul Owens (baseball)"
},
{
"operation": "finish",
"time": "354.602",
"value": "finish"
}
]
}
|
24866
|
Akysis vespa as described in 2004, the fish reaches over 10 centimeters and has a pattern of yellow-orange.
|
{
"evidence": [
"Akysis vespa_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:\nAkysis vespa as described in 2004, the fish reaches over 10 centimeters and has a pattern of yellow-orange.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Akysis vespa_sentence_3"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6675,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.189",
"value": "?search="
},
{
"operation": "search",
"time": "31.205",
"value": "Akysis vespa"
},
{
"operation": "Now on",
"time": "31.92",
"value": "Akysis vespa"
},
{
"operation": "Highlighting",
"time": "34.993",
"value": "Akysis vespa_sentence_2"
},
{
"operation": "Highlighting",
"time": "50.056",
"value": "Akysis vespa_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "50.735",
"value": "Akysis vespa_sentence_2"
},
{
"operation": "finish",
"time": "101.204",
"value": "finish"
}
]
}
|
exfever_2266
|
The Wolfman in 2010 is a remake of a horror film written by a German-American novelist and screenwriter.
|
{
"evidence": [
"The Wolfman (2010 film)",
"The Wolf Man (1941 film)",
"Curt Siodmak"
],
"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 Wolfman in 2010 is a remake of a horror film written by a German-American novelist and screenwriter.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Wolfman (2010 film)",
"The Wolf Man (1941 film)",
"Curt Siodmak"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6676,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_5307
|
21 Jump Street is a 2012 American buddy cop action comedy film directed by an American filmmaking duo that won an award that is given each year for animated films.
|
{
"evidence": [
"21 Jump Street (film)",
"Phil Lord and Christopher Miller",
"Academy Award for Best Animated Feature"
],
"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:\n21 Jump Street is a 2012 American buddy cop action comedy film directed by an American filmmaking duo that won an award that is given each year for animated films.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"21 Jump Street (film)",
"Phil Lord and Christopher Miller",
"Academy Award for Best Animated Feature"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6677,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
82768
|
Yeren means safe in Chinese.
|
{
"evidence": [
"Yeren_sentence_1",
"Yeren_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:\nYeren means safe in Chinese.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Yeren_sentence_1",
"Yeren_sentence_2"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6678,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.152",
"value": "?search="
},
{
"operation": "search",
"time": "9.601",
"value": "Yeren"
},
{
"operation": "Now on",
"time": "10.482",
"value": "Yeren"
},
{
"operation": "search",
"time": "32.012",
"value": "Yeren means safe in Chinese."
},
{
"operation": "Now on",
"time": "33.162",
"value": "undefined"
},
{
"operation": "back-button-clicked",
"time": "37.503",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "37.649",
"value": "Yeren"
},
{
"operation": "Highlighting",
"time": "77.879",
"value": "Yeren_sentence_3"
},
{
"operation": "Highlighting deleted",
"time": "78.527",
"value": "Yeren_sentence_3"
},
{
"operation": "hyperlink",
"time": "81.83",
"value": "Yeren (Zhou dynasty)"
},
{
"operation": "Now on",
"time": "82.344",
"value": "Yeren (Zhou dynasty)"
},
{
"operation": "back-button-clicked",
"time": "88.671",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "88.819",
"value": "Yeren"
},
{
"operation": "Highlighting",
"time": "98.815",
"value": "Yeren_sentence_1"
},
{
"operation": "Highlighting",
"time": "99.367",
"value": "Yeren_sentence_2"
},
{
"operation": "finish",
"time": "117.263",
"value": "finish"
}
]
}
|
exfever_1040
|
The Conjuring is a 2016 American supernatural horror film directed by an Australian director, producer, screenwriter and comic book writer who primarily worked in the genre of speculative fiction which is intended to frighten, scare, or disgust.
|
{
"evidence": [
"The Conjuring",
"James Wan",
"Horror fiction"
],
"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 Conjuring is a 2016 American supernatural horror film directed by an Australian director, producer, screenwriter and comic book writer who primarily worked in the genre of speculative fiction which is intended to frighten, scare, or disgust.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"The Conjuring",
"James Wan",
"Horror fiction"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6679,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
exfever_1393
|
Kingdom Hearts are the number one role-playing game in the world.
|
{
"evidence": [
"Kingdom Hearts",
"Square Enix"
],
"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:\nKingdom Hearts are the number one role-playing game in the world.\n",
"role": "user"
}
] |
EX-FEVER
|
{
"ground_truth": {
"evidence": [
"Kingdom Hearts",
"Square Enix"
],
"label": "NOT ENOUGH INFO"
},
"style": "rule"
}
|
{
"index": 6680,
"split": "train_online"
}
|
{
"annotator_operations": []
}
|
70617
|
Jenna Shoemaker, born on May 20, 1984, in Concord, Massachusetts, is an actor, writer, former professional US triathlete and was number 49 in the World Championship Series ranking.
|
{
"evidence": [
"Jenna Shoemaker_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:\nJenna Shoemaker, born on May 20, 1984, in Concord, Massachusetts, is an actor, writer, former professional US triathlete and was number 49 in the World Championship Series ranking.\n",
"role": "user"
}
] |
Other
|
{
"ground_truth": {
"evidence": [
"Jenna Shoemaker_sentence_0"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6681,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "0.714",
"value": "?search="
},
{
"operation": "search",
"time": "14.923",
"value": "Jenna Shoemaker"
},
{
"operation": "Now on",
"time": "17.065",
"value": "Jenna Shoemaker"
},
{
"operation": "Highlighting",
"time": "19.141",
"value": "Jenna Shoemaker_sentence_0"
},
{
"operation": "finish",
"time": "49.385",
"value": "finish"
}
]
}
|
50029
|
Phenylacetaldehyde dehydrogenase (participates in phenylalanine metabolism) in the family of oxidoreductases, is an enzyme that catalyzes the chemical reaction and it has around 2 substrates.
|
{
"evidence": [
"Phenylacetaldehyde 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:\nPhenylacetaldehyde dehydrogenase (participates in phenylalanine metabolism) in the family of oxidoreductases, is an enzyme that catalyzes the chemical reaction and it has around 2 substrates.\n",
"role": "user"
}
] |
Numerical Reasoning
|
{
"ground_truth": {
"evidence": [
"Phenylacetaldehyde dehydrogenase_sentence_1"
],
"label": "REFUTE"
},
"style": "rule"
}
|
{
"index": 6682,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.372",
"value": "?search="
},
{
"operation": "search",
"time": "9.556",
"value": "Phenylacetaldehyde dehydrogenase"
},
{
"operation": "Now on",
"time": "10.647",
"value": "Phenylacetaldehyde dehydrogenase"
},
{
"operation": "Page search",
"time": "32.064",
"value": "metabolism"
},
{
"operation": "Highlighting",
"time": "34.275",
"value": "Phenylacetaldehyde dehydrogenase_sentence_4"
},
{
"operation": "Highlighting",
"time": "48.952",
"value": "Phenylacetaldehyde dehydrogenase_sentence_2"
},
{
"operation": "hyperlink",
"time": "58.761",
"value": "oxidoreductases"
},
{
"operation": "Now on",
"time": "59.693",
"value": "Oxidoreductase"
},
{
"operation": "Highlighting",
"time": "80.491",
"value": "Oxidoreductase_sentence_0"
},
{
"operation": "Highlighting",
"time": "80.493",
"value": "Oxidoreductase_title"
},
{
"operation": "Page search",
"time": "105.442",
"value": "substrate"
},
{
"operation": "Highlighting",
"time": "120.563",
"value": "Oxidoreductase_sentence_1"
},
{
"operation": "hyperlink",
"time": "132.067",
"value": "NADP+"
},
{
"operation": "Now on",
"time": "133.476",
"value": "Nicotinamide adenine dinucleotide phosphate"
},
{
"operation": "Page search",
"time": "138.126",
"value": "substrate"
},
{
"operation": "hyperlink",
"time": "144.348",
"value": "4 Enzymes that use NADP(H) as a substrate Nicotinamide adenine dinucleotide phosphate_section_5"
},
{
"operation": "back-button-clicked",
"time": "155.742",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "156.014",
"value": "Oxidoreductase"
},
{
"operation": "Highlighting deleted",
"time": "163.03",
"value": "Oxidoreductase_sentence_1"
},
{
"operation": "back-button-clicked",
"time": "166.909",
"value": "back-button-clicked"
},
{
"operation": "back-button-clicked",
"time": "166.916",
"value": "back-button-clicked"
},
{
"operation": "Now on",
"time": "167.191",
"value": "Phenylacetaldehyde dehydrogenase"
},
{
"operation": "Page search",
"time": "174.842",
"value": "substrate"
},
{
"operation": "Highlighting",
"time": "178.016",
"value": "Phenylacetaldehyde dehydrogenase_sentence_1"
},
{
"operation": "Highlighting",
"time": "183.269",
"value": "Phenylacetaldehyde dehydrogenase_sentence_1"
},
{
"operation": "finish",
"time": "193.549",
"value": "finish"
}
]
}
|
92057
|
Miss Earth Uganda, which was created in 2014, is a beauty pageant in search of the most beautiful and environmentally-knowledgeable young Ugandan lady and is a good platform for the girls to get exposure and experience.
|
{
"evidence": [
"Miss Earth Uganda_sentence_0",
"Miss Earth Uganda_sentence_3",
"Miss Earth Uganda_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:\nMiss Earth Uganda, which was created in 2014, is a beauty pageant in search of the most beautiful and environmentally-knowledgeable young Ugandan lady and is a good platform for the girls to get exposure and experience.\n",
"role": "user"
}
] |
Multi-hop Reasoning
|
{
"ground_truth": {
"evidence": [
"Miss Earth Uganda_sentence_0",
"Miss Earth Uganda_sentence_3",
"Miss Earth Uganda_sentence_5"
],
"label": "SUPPORT"
},
"style": "rule"
}
|
{
"index": 6683,
"split": "train_online"
}
|
{
"annotator_operations": [
{
"operation": "start",
"time": "0",
"value": "start"
},
{
"operation": "Now on",
"time": "1.016",
"value": "?search="
},
{
"operation": "search",
"time": "28.155",
"value": "Miss Earth Uganda"
},
{
"operation": "Now on",
"time": "28.913",
"value": "Miss Earth Uganda"
},
{
"operation": "Highlighting",
"time": "30.754",
"value": "Miss Earth Uganda_sentence_0"
},
{
"operation": "Highlighting",
"time": "46.793",
"value": "Miss Earth Uganda_sentence_3"
},
{
"operation": "Highlighting",
"time": "48.217",
"value": "Miss Earth Uganda_sentence_5"
},
{
"operation": "finish",
"time": "56.169",
"value": "finish"
}
]
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.