data_source
stringclasses
2 values
prompt
listlengths
2
2
ability
stringclasses
1 value
reward_model
dict
extra_info
dict
metadata
dict
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: which indian captain has won the maximum matches against sri lanka?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Virat Kohli" ] }, "style": "rule" }
{ "index": 100, "need_tools_kwargs": true, "question": "which indian captain has won the maximum matches against sri lanka?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Virat Kohli" ] }, "question": "which indian captain has won the maximum matches against sri lanka?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is tide liquid called in the uk?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Daz Go-Pods" ] }, "style": "rule" }
{ "index": 101, "need_tools_kwargs": true, "question": "what is tide liquid called in the uk?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Daz Go-Pods" ] }, "question": "what is tide liquid called in the uk?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what's the dog's name on garfield?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Odie" ] }, "style": "rule" }
{ "index": 102, "need_tools_kwargs": true, "question": "what's the dog's name on garfield?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Odie" ] }, "question": "what's the dog's name on garfield?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: five australian prime ministers who governed during world war 2?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Ben Chifley", "Frank Forde", "John Curtin", "Arthur Fadden", "Robert Menzies" ] }, "style": "rule" }
{ "index": 103, "need_tools_kwargs": true, "question": "five australian prime ministers who governed during world war 2?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Ben Chifley", "Frank Forde", "John Curtin", "Arthur Fadden", "Robert Menzies" ] }, "question": "five australian prime ministers who governed during world war 2?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the meaning of the last name salazar?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "old hall" ] }, "style": "rule" }
{ "index": 104, "need_tools_kwargs": true, "question": "what is the meaning of the last name salazar?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "old hall" ] }, "question": "what is the meaning of the last name salazar?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who was a in pretty little liars episode?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "CeCe Drake", "Mona Vanderwaal", "Alex Drake", "Sara Harvey" ] }, "style": "rule" }
{ "index": 105, "need_tools_kwargs": true, "question": "who was a in pretty little liars episode?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "CeCe Drake", "Mona Vanderwaal", "Alex Drake", "Sara Harvey" ] }, "question": "who was a in pretty little liars episode?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: qualification of the members of the supreme court in the philippines?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "natural-born citizenship", "at least 40 years old" ] }, "style": "rule" }
{ "index": 106, "need_tools_kwargs": true, "question": "qualification of the members of the supreme court in the philippines?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "natural-born citizenship", "at least 40 years old" ] }, "question": "qualification of the members of the supreme court in the philippines?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays finn on my mad fat diary?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Nico Cristian Mirallegro" ] }, "style": "rule" }
{ "index": 107, "need_tools_kwargs": true, "question": "who plays finn on my mad fat diary?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Nico Cristian Mirallegro" ] }, "question": "who plays finn on my mad fat diary?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: which type of technology encapsulates fibre channel communications over tcp/ip?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "iSCSI" ] }, "style": "rule" }
{ "index": 108, "need_tools_kwargs": true, "question": "which type of technology encapsulates fibre channel communications over tcp/ip?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "iSCSI" ] }, "question": "which type of technology encapsulates fibre channel communications over tcp/ip?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who was president during attack on pearl harbor?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "President Franklin D. Roosevelt" ] }, "style": "rule" }
{ "index": 109, "need_tools_kwargs": true, "question": "who was president during attack on pearl harbor?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "President Franklin D. Roosevelt" ] }, "question": "who was president during attack on pearl harbor?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where does the umbilical vein enter the circulation of the fetus?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the portal vein" ] }, "style": "rule" }
{ "index": 110, "need_tools_kwargs": true, "question": "where does the umbilical vein enter the circulation of the fetus?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the portal vein" ] }, "question": "where does the umbilical vein enter the circulation of the fetus?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did def leppard's drummer lose his arm?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1985" ] }, "style": "rule" }
{ "index": 111, "need_tools_kwargs": true, "question": "when did def leppard's drummer lose his arm?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1985" ] }, "question": "when did def leppard's drummer lose his arm?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: whose portrait appears on the choir call to the right of the apse of san vitale ravenna?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "East Roman Emperor Justinian I" ] }, "style": "rule" }
{ "index": 112, "need_tools_kwargs": true, "question": "whose portrait appears on the choir call to the right of the apse of san vitale ravenna?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "East Roman Emperor Justinian I" ] }, "question": "whose portrait appears on the choir call to the right of the apse of san vitale ravenna?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays stephanie's mom on the fosters?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Annie Potts" ] }, "style": "rule" }
{ "index": 113, "need_tools_kwargs": true, "question": "who plays stephanie's mom on the fosters?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Annie Potts" ] }, "question": "who plays stephanie's mom on the fosters?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: the unit of area in mks system is?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "metre" ] }, "style": "rule" }
{ "index": 114, "need_tools_kwargs": true, "question": "the unit of area in mks system is?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "metre" ] }, "question": "the unit of area in mks system is?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the name of the college in animal house?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Faber College" ] }, "style": "rule" }
{ "index": 115, "need_tools_kwargs": true, "question": "what is the name of the college in animal house?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Faber College" ] }, "question": "what is the name of the college in animal house?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what movie is the song magic carpet ride in?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Reservoir Dogs", "Coneheads" ] }, "style": "rule" }
{ "index": 116, "need_tools_kwargs": true, "question": "what movie is the song magic carpet ride in?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Reservoir Dogs", "Coneheads" ] }, "question": "what movie is the song magic carpet ride in?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who owned the louisiana territory when the constitution was signed?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Spain" ] }, "style": "rule" }
{ "index": 117, "need_tools_kwargs": true, "question": "who owned the louisiana territory when the constitution was signed?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Spain" ] }, "question": "who owned the louisiana territory when the constitution was signed?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when is the last time the pirates won the world series?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1979" ] }, "style": "rule" }
{ "index": 118, "need_tools_kwargs": true, "question": "when is the last time the pirates won the world series?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1979" ] }, "question": "when is the last time the pirates won the world series?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what's the name of the little turtle in finding nemo?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Squirt" ] }, "style": "rule" }
{ "index": 119, "need_tools_kwargs": true, "question": "what's the name of the little turtle in finding nemo?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Squirt" ] }, "question": "what's the name of the little turtle in finding nemo?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays the voice of belle in beauty and the beast?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Paige O'Hara" ] }, "style": "rule" }
{ "index": 120, "need_tools_kwargs": true, "question": "who plays the voice of belle in beauty and the beast?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Paige O'Hara" ] }, "question": "who plays the voice of belle in beauty and the beast?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was the hero with a thousand faces published?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1949" ] }, "style": "rule" }
{ "index": 121, "need_tools_kwargs": true, "question": "when was the hero with a thousand faces published?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1949" ] }, "question": "when was the hero with a thousand faces published?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did the united states begin its revolution?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "April 1775" ] }, "style": "rule" }
{ "index": 122, "need_tools_kwargs": true, "question": "when did the united states begin its revolution?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "April 1775" ] }, "question": "when did the united states begin its revolution?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played cosette in les miserables 25th anniversary?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Katie Hall" ] }, "style": "rule" }
{ "index": 123, "need_tools_kwargs": true, "question": "who played cosette in les miserables 25th anniversary?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Katie Hall" ] }, "question": "who played cosette in les miserables 25th anniversary?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was the last time the falcons were in the super bowl?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "2016" ] }, "style": "rule" }
{ "index": 124, "need_tools_kwargs": true, "question": "when was the last time the falcons were in the super bowl?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "2016" ] }, "question": "when was the last time the falcons were in the super bowl?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the default administrative distance assigned to all ospf routes?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "110" ] }, "style": "rule" }
{ "index": 125, "need_tools_kwargs": true, "question": "what is the default administrative distance assigned to all ospf routes?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "110" ] }, "question": "what is the default administrative distance assigned to all ospf routes?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who heads the u.s. intelligence community science forum?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Richard N. Haass" ] }, "style": "rule" }
{ "index": 126, "need_tools_kwargs": true, "question": "who heads the u.s. intelligence community science forum?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Richard N. Haass" ] }, "question": "who heads the u.s. intelligence community science forum?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who is the most decorated veteran in us history?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Audie Leon Murphy" ] }, "style": "rule" }
{ "index": 127, "need_tools_kwargs": true, "question": "who is the most decorated veteran in us history?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Audie Leon Murphy" ] }, "question": "who is the most decorated veteran in us history?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who has more la liga titles real madrid or barcelona?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Real Madrid" ] }, "style": "rule" }
{ "index": 128, "need_tools_kwargs": true, "question": "who has more la liga titles real madrid or barcelona?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Real Madrid" ] }, "question": "who has more la liga titles real madrid or barcelona?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who is ryan michelle bathe on this is us?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Yvette" ] }, "style": "rule" }
{ "index": 129, "need_tools_kwargs": true, "question": "who is ryan michelle bathe on this is us?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Yvette" ] }, "question": "who is ryan michelle bathe on this is us?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who burned down the billboards in three billboards?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Charlie" ] }, "style": "rule" }
{ "index": 130, "need_tools_kwargs": true, "question": "who burned down the billboards in three billboards?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Charlie" ] }, "question": "who burned down the billboards in three billboards?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when does the supreme court begin a new term?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "first Monday of each October" ] }, "style": "rule" }
{ "index": 131, "need_tools_kwargs": true, "question": "when does the supreme court begin a new term?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "first Monday of each October" ] }, "question": "when does the supreme court begin a new term?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when does dragon ball z super episode 105 come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "August 27, 2017" ] }, "style": "rule" }
{ "index": 132, "need_tools_kwargs": true, "question": "when does dragon ball z super episode 105 come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "August 27, 2017" ] }, "question": "when does dragon ball z super episode 105 come out?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played dean on wizards of waverly place?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Samonas" ] }, "style": "rule" }
{ "index": 133, "need_tools_kwargs": true, "question": "who played dean on wizards of waverly place?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Samonas" ] }, "question": "who played dean on wizards of waverly place?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who sings i just died in your arms?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Cutting Crew" ] }, "style": "rule" }
{ "index": 134, "need_tools_kwargs": true, "question": "who sings i just died in your arms?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Cutting Crew" ] }, "question": "who sings i just died in your arms?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the frequency of ac current in north america?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "60 Hz" ] }, "style": "rule" }
{ "index": 135, "need_tools_kwargs": true, "question": "what is the frequency of ac current in north america?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "60 Hz" ] }, "question": "what is the frequency of ac current in north america?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: in which episode does goku arrive on namek?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "65" ] }, "style": "rule" }
{ "index": 136, "need_tools_kwargs": true, "question": "in which episode does goku arrive on namek?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "65" ] }, "question": "in which episode does goku arrive on namek?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did zebra mussels come to north america?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1988" ] }, "style": "rule" }
{ "index": 137, "need_tools_kwargs": true, "question": "when did zebra mussels come to north america?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1988" ] }, "question": "when did zebra mussels come to north america?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays the voice of tiana in princess and the frog?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Anika Noni Rose" ] }, "style": "rule" }
{ "index": 138, "need_tools_kwargs": true, "question": "who plays the voice of tiana in princess and the frog?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Anika Noni Rose" ] }, "question": "who plays the voice of tiana in princess and the frog?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who developed a method of acting to reproduce real human behavior on the stage?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Stanislavski" ] }, "style": "rule" }
{ "index": 139, "need_tools_kwargs": true, "question": "who developed a method of acting to reproduce real human behavior on the stage?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Stanislavski" ] }, "question": "who developed a method of acting to reproduce real human behavior on the stage?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did they come out with color photos?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1861" ] }, "style": "rule" }
{ "index": 140, "need_tools_kwargs": true, "question": "when did they come out with color photos?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1861" ] }, "question": "when did they come out with color photos?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who made call of duty world at war?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Treyarch" ] }, "style": "rule" }
{ "index": 141, "need_tools_kwargs": true, "question": "who made call of duty world at war?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Treyarch" ] }, "question": "who made call of duty world at war?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where in london is call the midwife set?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the East End" ] }, "style": "rule" }
{ "index": 142, "need_tools_kwargs": true, "question": "where in london is call the midwife set?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the East End" ] }, "question": "where in london is call the midwife set?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: the organization of the formal elements in an art work?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "composition" ] }, "style": "rule" }
{ "index": 143, "need_tools_kwargs": true, "question": "the organization of the formal elements in an art work?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "composition" ] }, "question": "the organization of the formal elements in an art work?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who wrote i wanna grow old with you?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Robbie" ] }, "style": "rule" }
{ "index": 144, "need_tools_kwargs": true, "question": "who wrote i wanna grow old with you?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Robbie" ] }, "question": "who wrote i wanna grow old with you?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did 10 rupees coin launched in india?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "2005" ] }, "style": "rule" }
{ "index": 145, "need_tools_kwargs": true, "question": "when did 10 rupees coin launched in india?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "2005" ] }, "question": "when did 10 rupees coin launched in india?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when does the third maze runner come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "January 26, 2018" ] }, "style": "rule" }
{ "index": 146, "need_tools_kwargs": true, "question": "when does the third maze runner come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "January 26, 2018" ] }, "question": "when does the third maze runner come out?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who dies in season 6 once upon a time?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Hyde" ] }, "style": "rule" }
{ "index": 147, "need_tools_kwargs": true, "question": "who dies in season 6 once upon a time?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Hyde" ] }, "question": "who dies in season 6 once upon a time?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where did the idea of retirement come from?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the 18th century" ] }, "style": "rule" }
{ "index": 148, "need_tools_kwargs": true, "question": "where did the idea of retirement come from?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the 18th century" ] }, "question": "where did the idea of retirement come from?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who's played the most years in the nba?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Robert Parish", "Kevin Garnett", "Kevin Willis" ] }, "style": "rule" }
{ "index": 149, "need_tools_kwargs": true, "question": "who's played the most years in the nba?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Robert Parish", "Kevin Garnett", "Kevin Willis" ] }, "question": "who's played the most years in the nba?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: how many prime ministers in the queen's reign?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "160" ] }, "style": "rule" }
{ "index": 150, "need_tools_kwargs": true, "question": "how many prime ministers in the queen's reign?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "160" ] }, "question": "how many prime ministers in the queen's reign?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who starred in the secret life of bees?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Paul Bettany", "Hilarie Burton", "Dakota Fanning", "Sophie Okonedo", "Queen Latifah", "Alicia Keys", "Jennifer Hudson" ] }, "style": "rule" }
{ "index": 151, "need_tools_kwargs": true, "question": "who starred in the secret life of bees?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Paul Bettany", "Hilarie Burton", "Dakota Fanning", "Sophie Okonedo", "Queen Latifah", "Alicia Keys", "Jennifer Hudson" ] }, "question": "who starred in the secret life of bees?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did the first macbook pro retina come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "October 23, 2012" ] }, "style": "rule" }
{ "index": 152, "need_tools_kwargs": true, "question": "when did the first macbook pro retina come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "October 23, 2012" ] }, "question": "when did the first macbook pro retina come out?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who won the women's wimbledon title in 2018?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Angelique Kerber" ] }, "style": "rule" }
{ "index": 153, "need_tools_kwargs": true, "question": "who won the women's wimbledon title in 2018?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Angelique Kerber" ] }, "question": "who won the women's wimbledon title in 2018?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where do you hit to test your reflexes?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the tendon in question" ] }, "style": "rule" }
{ "index": 154, "need_tools_kwargs": true, "question": "where do you hit to test your reflexes?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the tendon in question" ] }, "question": "where do you hit to test your reflexes?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who was the king of england in the second world war?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "George VI" ] }, "style": "rule" }
{ "index": 155, "need_tools_kwargs": true, "question": "who was the king of england in the second world war?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "George VI" ] }, "question": "who was the king of england in the second world war?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: three types of native mammals that live in florida?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Carnivora", "Rodentia", "Cetacea" ] }, "style": "rule" }
{ "index": 156, "need_tools_kwargs": true, "question": "three types of native mammals that live in florida?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Carnivora", "Rodentia", "Cetacea" ] }, "question": "three types of native mammals that live in florida?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who was named time person of the year?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "The Silence Breakers" ] }, "style": "rule" }
{ "index": 157, "need_tools_kwargs": true, "question": "who was named time person of the year?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "The Silence Breakers" ] }, "question": "who was named time person of the year?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where was the aztec civilization located on a map?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "central Mexico" ] }, "style": "rule" }
{ "index": 158, "need_tools_kwargs": true, "question": "where was the aztec civilization located on a map?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "central Mexico" ] }, "question": "where was the aztec civilization located on a map?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was the seat belt law introduced in australia?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1971", "1970", "By 1973" ] }, "style": "rule" }
{ "index": 159, "need_tools_kwargs": true, "question": "when was the seat belt law introduced in australia?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1971", "1970", "By 1973" ] }, "question": "when was the seat belt law introduced in australia?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when does the mlb regular season start in 2018?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "March 29, 2018" ] }, "style": "rule" }
{ "index": 160, "need_tools_kwargs": true, "question": "when does the mlb regular season start in 2018?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "March 29, 2018" ] }, "question": "when does the mlb regular season start in 2018?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what is the name of the russian fur hat?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "An ushanka" ] }, "style": "rule" }
{ "index": 161, "need_tools_kwargs": true, "question": "what is the name of the russian fur hat?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "An ushanka" ] }, "question": "what is the name of the russian fur hat?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: types of rnas carries information for making a single type of protein?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "messenger RNA (mRNA)" ] }, "style": "rule" }
{ "index": 162, "need_tools_kwargs": true, "question": "types of rnas carries information for making a single type of protein?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "messenger RNA (mRNA)" ] }, "question": "types of rnas carries information for making a single type of protein?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was ain't put into the dictionary?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1961" ] }, "style": "rule" }
{ "index": 163, "need_tools_kwargs": true, "question": "when was ain't put into the dictionary?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1961" ] }, "question": "when was ain't put into the dictionary?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who coined the term protoplasm for the fluid substance of the cell?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "J. E. Purkinje" ] }, "style": "rule" }
{ "index": 164, "need_tools_kwargs": true, "question": "who coined the term protoplasm for the fluid substance of the cell?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "J. E. Purkinje" ] }, "question": "who coined the term protoplasm for the fluid substance of the cell?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: how has the most mvp in nba history?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Kareem Abdul-Jabbar" ] }, "style": "rule" }
{ "index": 165, "need_tools_kwargs": true, "question": "how has the most mvp in nba history?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Kareem Abdul-Jabbar" ] }, "question": "how has the most mvp in nba history?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when is the show this is us on tv?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "9:00 pm", "Tuesday" ] }, "style": "rule" }
{ "index": 166, "need_tools_kwargs": true, "question": "when is the show this is us on tv?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "9:00 pm", "Tuesday" ] }, "question": "when is the show this is us on tv?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where was the first petro-chemical industry established in nigeria?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1907" ] }, "style": "rule" }
{ "index": 167, "need_tools_kwargs": true, "question": "where was the first petro-chemical industry established in nigeria?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1907" ] }, "question": "where was the first petro-chemical industry established in nigeria?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who does sarah chalke voice in rick and morty?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Beth Smith" ] }, "style": "rule" }
{ "index": 168, "need_tools_kwargs": true, "question": "who does sarah chalke voice in rick and morty?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Beth Smith" ] }, "question": "who does sarah chalke voice in rick and morty?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who is responsible for managing medicaid insurance benefits?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "government", "the United States" ] }, "style": "rule" }
{ "index": 169, "need_tools_kwargs": true, "question": "who is responsible for managing medicaid insurance benefits?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "government", "the United States" ] }, "question": "who is responsible for managing medicaid insurance benefits?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays jt in the young and the restless?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Thaddeus Rowe Luckinbill" ] }, "style": "rule" }
{ "index": 170, "need_tools_kwargs": true, "question": "who plays jt in the young and the restless?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Thaddeus Rowe Luckinbill" ] }, "question": "who plays jt in the young and the restless?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played buffalo hump in dead man's walk?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Eric Schweig" ] }, "style": "rule" }
{ "index": 171, "need_tools_kwargs": true, "question": "who played buffalo hump in dead man's walk?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Eric Schweig" ] }, "question": "who played buffalo hump in dead man's walk?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was the last episode of everyone loves raymond?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "May 16, 2005" ] }, "style": "rule" }
{ "index": 172, "need_tools_kwargs": true, "question": "when was the last episode of everyone loves raymond?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "May 16, 2005" ] }, "question": "when was the last episode of everyone loves raymond?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where does glycosis take place in the cell?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "cytosol" ] }, "style": "rule" }
{ "index": 173, "need_tools_kwargs": true, "question": "where does glycosis take place in the cell?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "cytosol" ] }, "question": "where does glycosis take place in the cell?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who wrote the hymn i need thee every hour?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Annie Sherwood Hawks" ] }, "style": "rule" }
{ "index": 174, "need_tools_kwargs": true, "question": "who wrote the hymn i need thee every hour?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Annie Sherwood Hawks" ] }, "question": "who wrote the hymn i need thee every hour?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays belle in beauty and the beast movie?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Emma Watson" ] }, "style": "rule" }
{ "index": 175, "need_tools_kwargs": true, "question": "who plays belle in beauty and the beast movie?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Emma Watson" ] }, "question": "who plays belle in beauty and the beast movie?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays poussey on orange is the new black?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Samira Wiley" ] }, "style": "rule" }
{ "index": 176, "need_tools_kwargs": true, "question": "who plays poussey on orange is the new black?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Samira Wiley" ] }, "question": "who plays poussey on orange is the new black?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: in javanese gamelan music typical elaborating instruments are?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "percussive" ] }, "style": "rule" }
{ "index": 177, "need_tools_kwargs": true, "question": "in javanese gamelan music typical elaborating instruments are?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "percussive" ] }, "question": "in javanese gamelan music typical elaborating instruments are?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who is the main character in memoirs of a geisha?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Chiyo Sakamoto", "\"Sayuri\"" ] }, "style": "rule" }
{ "index": 178, "need_tools_kwargs": true, "question": "who is the main character in memoirs of a geisha?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Chiyo Sakamoto", "\"Sayuri\"" ] }, "question": "who is the main character in memoirs of a geisha?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: halloween the curse of michael myers paul rudd?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Tommy Doyle" ] }, "style": "rule" }
{ "index": 179, "need_tools_kwargs": true, "question": "halloween the curse of michael myers paul rudd?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Tommy Doyle" ] }, "question": "halloween the curse of michael myers paul rudd?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played jonathan in arsenic and old lace?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Raymond Massey", "Boris Karloff" ] }, "style": "rule" }
{ "index": 180, "need_tools_kwargs": true, "question": "who played jonathan in arsenic and old lace?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Raymond Massey", "Boris Karloff" ] }, "question": "who played jonathan in arsenic and old lace?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who ran the fastest 40 yard dash in history?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jakeem Grant" ] }, "style": "rule" }
{ "index": 181, "need_tools_kwargs": true, "question": "who ran the fastest 40 yard dash in history?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jakeem Grant" ] }, "question": "who ran the fastest 40 yard dash in history?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did budapest became the capital of hungary?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1361" ] }, "style": "rule" }
{ "index": 182, "need_tools_kwargs": true, "question": "when did budapest became the capital of hungary?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1361" ] }, "question": "when did budapest became the capital of hungary?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: where is the first place to bring in the new year?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Tonga", "Samoa" ] }, "style": "rule" }
{ "index": 183, "need_tools_kwargs": true, "question": "where is the first place to bring in the new year?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Tonga", "Samoa" ] }, "question": "where is the first place to bring in the new year?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played ariel in the little mermaid on broadway?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Sierra Boggess" ] }, "style": "rule" }
{ "index": 184, "need_tools_kwargs": true, "question": "who played ariel in the little mermaid on broadway?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Sierra Boggess" ] }, "question": "who played ariel in the little mermaid on broadway?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what part of war and peace is the great comet?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Volume 2, Part 5" ] }, "style": "rule" }
{ "index": 185, "need_tools_kwargs": true, "question": "what part of war and peace is the great comet?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Volume 2, Part 5" ] }, "question": "what part of war and peace is the great comet?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what was the first attack in world war 2?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Invasion of Poland" ] }, "style": "rule" }
{ "index": 186, "need_tools_kwargs": true, "question": "what was the first attack in world war 2?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Invasion of Poland" ] }, "question": "what was the first attack in world war 2?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did mardi gras start in new orleans?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "March 2, 1699" ] }, "style": "rule" }
{ "index": 187, "need_tools_kwargs": true, "question": "when did mardi gras start in new orleans?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "March 2, 1699" ] }, "question": "when did mardi gras start in new orleans?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who won best actor academy award in 1969?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Cliff Robertson" ] }, "style": "rule" }
{ "index": 188, "need_tools_kwargs": true, "question": "who won best actor academy award in 1969?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Cliff Robertson" ] }, "question": "who won best actor academy award in 1969?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who plays the doctor in sons of anarchy?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Siff" ] }, "style": "rule" }
{ "index": 189, "need_tools_kwargs": true, "question": "who plays the doctor in sons of anarchy?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Siff" ] }, "question": "who plays the doctor in sons of anarchy?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who wrote oh what a tangled web we weave when first we practice to deceive?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Canto VI, XVII" ] }, "style": "rule" }
{ "index": 190, "need_tools_kwargs": true, "question": "who wrote oh what a tangled web we weave when first we practice to deceive?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Canto VI, XVII" ] }, "question": "who wrote oh what a tangled web we weave when first we practice to deceive?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when did disney world start using magic bands?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "2013" ] }, "style": "rule" }
{ "index": 191, "need_tools_kwargs": true, "question": "when did disney world start using magic bands?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "2013" ] }, "question": "when did disney world start using magic bands?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who played the brothers in seven brides for seven brothers?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jeff Richards", "Tommy Rall", "Matt Mattox", "Marc Platt", "Jacques d'Amboise", "Howard Keel", "Russ Tamblyn" ] }, "style": "rule" }
{ "index": 192, "need_tools_kwargs": true, "question": "who played the brothers in seven brides for seven brothers?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jeff Richards", "Tommy Rall", "Matt Mattox", "Marc Platt", "Jacques d'Amboise", "Howard Keel", "Russ Tamblyn" ] }, "question": "who played the brothers in seven brides for seven brothers?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who sings the theme song to dog bounty hunter?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Ozzy Osbourne" ] }, "style": "rule" }
{ "index": 193, "need_tools_kwargs": true, "question": "who sings the theme song to dog bounty hunter?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Ozzy Osbourne" ] }, "question": "who sings the theme song to dog bounty hunter?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when is the last time the minnesota vikings have been in the playoffs?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "2017" ] }, "style": "rule" }
{ "index": 194, "need_tools_kwargs": true, "question": "when is the last time the minnesota vikings have been in the playoffs?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "2017" ] }, "question": "when is the last time the minnesota vikings have been in the playoffs?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: when was the qutub minar granted the world heritage site status?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1993" ] }, "style": "rule" }
{ "index": 195, "need_tools_kwargs": true, "question": "when was the qutub minar granted the world heritage site status?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1993" ] }, "question": "when was the qutub minar granted the world heritage site status?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who was the first civilian governor of lagos state?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Alhaji Lateef Jakande" ] }, "style": "rule" }
{ "index": 196, "need_tools_kwargs": true, "question": "who was the first civilian governor of lagos state?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Alhaji Lateef Jakande" ] }, "question": "who was the first civilian governor of lagos state?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who is first on record as proposing a geocentric universe?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Anaximander" ] }, "style": "rule" }
{ "index": 197, "need_tools_kwargs": true, "question": "who is first on record as proposing a geocentric universe?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Anaximander" ] }, "question": "who is first on record as proposing a geocentric universe?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: what do they call high school in australia?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "(high schools" ] }, "style": "rule" }
{ "index": 198, "need_tools_kwargs": true, "question": "what do they call high school in australia?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "(high schools" ] }, "question": "what do they call high school in australia?" } } } }
null
nq
[ { "content": "You are a helpful and harmless assistant.", "role": "system" }, { "content": "Answer the given question step by step.\n Instructions:\n 1. First, conduct reasoning inside <think> and </think> tags whenever you receive new information.\n 2. If you need external knowledge, you can search using <search>query</search>.\n 3. When you receive search results, evaluate their usefulness and put your judgment inside <judge>Yes</judge> or <judge>No</judge> tags.\n 4. Based on your judgment, follow these strict rules:\n a. If the information is useful AND you now have sufficient information to provide a complete final answer, proceed directly to step 5.\n b. If the information is useful BUT you still need more details, you MUST search again with <search>...</search>.\n c. If the information is not useful, you MUST search again with <search>...</search>. You MUST NOT provide an answer in this case.\n 5. Provide your final answer in <answer>...</answer> tags. The <answer> tag marks the end of the task. After providing the <answer>, you MUST stop and generate no further text.\n\n Question: who does young catherine marry in wuthering heights?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Hareton Earnshaw" ] }, "style": "rule" }
{ "index": 199, "need_tools_kwargs": true, "question": "who does young catherine marry in wuthering heights?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Hareton Earnshaw" ] }, "question": "who does young catherine marry in wuthering heights?" } } } }
null