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: when did spider man first appear in comics?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "August 1962" ] }, "style": "rule" }
{ "index": 200, "need_tools_kwargs": true, "question": "when did spider man first appear in comics?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "August 1962" ] }, "question": "when did spider man first appear in comics?" } } } }
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 they go in the inbetweeners movie?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Malia" ] }, "style": "rule" }
{ "index": 201, "need_tools_kwargs": true, "question": "where did they go in the inbetweeners movie?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Malia" ] }, "question": "where did they go in the inbetweeners 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 sings if you want to destroy my sweater?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Weezer" ] }, "style": "rule" }
{ "index": 202, "need_tools_kwargs": true, "question": "who sings if you want to destroy my sweater?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Weezer" ] }, "question": "who sings if you want to destroy my sweater?" } } } }
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 bernadette in the big bang theory?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Melissa Ivy Rauch" ] }, "style": "rule" }
{ "index": 203, "need_tools_kwargs": true, "question": "who plays bernadette in the big bang theory?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Melissa Ivy Rauch" ] }, "question": "who plays bernadette in the big bang theory?" } } } }
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 mirabell's servant?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Waitwell" ] }, "style": "rule" }
{ "index": 204, "need_tools_kwargs": true, "question": "what is the name of mirabell's servant?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Waitwell" ] }, "question": "what is the name of mirabell's servant?" } } } }
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 france hosted the olympics?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1992 Winter Olympics" ] }, "style": "rule" }
{ "index": 205, "need_tools_kwargs": true, "question": "when was the last time france hosted the olympics?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1992 Winter Olympics" ] }, "question": "when was the last time france hosted the olympics?" } } } }
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 sang too shy shy hush hush eye to eye?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Kajagoogoo" ] }, "style": "rule" }
{ "index": 206, "need_tools_kwargs": true, "question": "who sang too shy shy hush hush eye to eye?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Kajagoogoo" ] }, "question": "who sang too shy shy hush hush eye to eye?" } } } }
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 first session of indian national congress held?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "28–31 December 1885" ] }, "style": "rule" }
{ "index": 207, "need_tools_kwargs": true, "question": "when was the first session of indian national congress held?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "28–31 December 1885" ] }, "question": "when was the first session of indian national congress held?" } } } }
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: for a molecule to be polar it must have?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "polar bonds" ] }, "style": "rule" }
{ "index": 208, "need_tools_kwargs": true, "question": "for a molecule to be polar it must have?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "polar bonds" ] }, "question": "for a molecule to be polar it must have?" } } } }
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 l.o.v.e in the movie the parent trap?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Nat King Cole" ] }, "style": "rule" }
{ "index": 209, "need_tools_kwargs": true, "question": "who sings l.o.v.e in the movie the parent trap?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Nat King Cole" ] }, "question": "who sings l.o.v.e in the movie the parent trap?" } } } }
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 rationing start in britain during world war 2?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1940" ] }, "style": "rule" }
{ "index": 210, "need_tools_kwargs": true, "question": "when did rationing start in britain during world war 2?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1940" ] }, "question": "when did rationing start in britain 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: who sings system from queen of the damned?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Chester Bennington of Linkin Park" ] }, "style": "rule" }
{ "index": 211, "need_tools_kwargs": true, "question": "who sings system from queen of the damned?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Chester Bennington of Linkin Park" ] }, "question": "who sings system from queen of the damned?" } } } }
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 ontario get rid of grade 13?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1988" ] }, "style": "rule" }
{ "index": 212, "need_tools_kwargs": true, "question": "when did ontario get rid of grade 13?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1988" ] }, "question": "when did ontario get rid of grade 13?" } } } }
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 first friday the 13th movie released?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1980" ] }, "style": "rule" }
{ "index": 213, "need_tools_kwargs": true, "question": "when was the first friday the 13th movie released?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1980" ] }, "question": "when was the first friday the 13th movie released?" } } } }
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 kirk's mother in star trek?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jennifer Marie Morrison" ] }, "style": "rule" }
{ "index": 214, "need_tools_kwargs": true, "question": "who played kirk's mother in star trek?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jennifer Marie Morrison" ] }, "question": "who played kirk's mother in star trek?" } } } }
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 how the grinch stole christmas written?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "beginning of 1957" ] }, "style": "rule" }
{ "index": 215, "need_tools_kwargs": true, "question": "when was how the grinch stole christmas written?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "beginning of 1957" ] }, "question": "when was how the grinch stole christmas written?" } } } }
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: why do they call easter island easter island?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "encountered it on Easter Sunday" ] }, "style": "rule" }
{ "index": 216, "need_tools_kwargs": true, "question": "why do they call easter island easter island?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "encountered it on Easter Sunday" ] }, "question": "why do they call easter island easter island?" } } } }
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 constitutes the finance commission after every five years?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "The President" ] }, "style": "rule" }
{ "index": 217, "need_tools_kwargs": true, "question": "who constitutes the finance commission after every five years?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "The President" ] }, "question": "who constitutes the finance commission after every five years?" } } } }
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 she's a bad mamma jamma?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Leon Haywood" ] }, "style": "rule" }
{ "index": 218, "need_tools_kwargs": true, "question": "who wrote she's a bad mamma jamma?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Leon Haywood" ] }, "question": "who wrote she's a bad mamma jamma?" } } } }
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 votes for winners of los premios juventud?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "online vote" ] }, "style": "rule" }
{ "index": 219, "need_tools_kwargs": true, "question": "who votes for winners of los premios juventud?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "online vote" ] }, "question": "who votes for winners of los premios juventud?" } } } }
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 happened to many manuscripts of anglo-saxon literature?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the Dissolution of the Monasteries" ] }, "style": "rule" }
{ "index": 220, "need_tools_kwargs": true, "question": "what happened to many manuscripts of anglo-saxon literature?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the Dissolution of the Monasteries" ] }, "question": "what happened to many manuscripts of anglo-saxon literature?" } } } }
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 are most nutrients absorbed in the small intestine?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "in the jejunum" ] }, "style": "rule" }
{ "index": 221, "need_tools_kwargs": true, "question": "where are most nutrients absorbed in the small intestine?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "in the jejunum" ] }, "question": "where are most nutrients absorbed in the small intestine?" } } } }
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 awarded with the first param veer chakra vijeta?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Major Somnath Sharma" ] }, "style": "rule" }
{ "index": 222, "need_tools_kwargs": true, "question": "who was awarded with the first param veer chakra vijeta?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Major Somnath Sharma" ] }, "question": "who was awarded with the first param veer chakra vijeta?" } } } }
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 tom lehrer write the elements song?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1959" ] }, "style": "rule" }
{ "index": 223, "need_tools_kwargs": true, "question": "when did tom lehrer write the elements song?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1959" ] }, "question": "when did tom lehrer write the elements song?" } } } }
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 doing the halftime show for the super bowl?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Justin Timberlake" ] }, "style": "rule" }
{ "index": 224, "need_tools_kwargs": true, "question": "who is doing the halftime show for the super bowl?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Justin Timberlake" ] }, "question": "who is doing the halftime show for 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: when's the last time the dodgers were in the world series?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "2017" ] }, "style": "rule" }
{ "index": 225, "need_tools_kwargs": true, "question": "when's the last time the dodgers were in the world series?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "2017" ] }, "question": "when's the last time the dodgers were in 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: who sang some enchanted evening in the film south pacific?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Ezio Pinza" ] }, "style": "rule" }
{ "index": 226, "need_tools_kwargs": true, "question": "who sang some enchanted evening in the film south pacific?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Ezio Pinza" ] }, "question": "who sang some enchanted evening in the film south pacific?" } } } }
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 diary of a wimpy kid the long haul come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "May 19, 2017" ] }, "style": "rule" }
{ "index": 227, "need_tools_kwargs": true, "question": "when did diary of a wimpy kid the long haul come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "May 19, 2017" ] }, "question": "when did diary of a wimpy kid the long haul 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: what scientists is credited with proposing the principle of uniformitarianism?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "James Hutton" ] }, "style": "rule" }
{ "index": 228, "need_tools_kwargs": true, "question": "what scientists is credited with proposing the principle of uniformitarianism?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "James Hutton" ] }, "question": "what scientists is credited with proposing the principle of uniformitarianism?" } } } }
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 vocals on go your own way?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Christine McVie", "Lindsey Buckingham", "Stevie Nicks" ] }, "style": "rule" }
{ "index": 229, "need_tools_kwargs": true, "question": "who sings vocals on go your own way?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Christine McVie", "Lindsey Buckingham", "Stevie Nicks" ] }, "question": "who sings vocals on go your own way?" } } } }
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 channel is paramount tv on direct tv?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "241 SD/HD", "1241 VOD" ] }, "style": "rule" }
{ "index": 230, "need_tools_kwargs": true, "question": "what channel is paramount tv on direct tv?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "241 SD/HD", "1241 VOD" ] }, "question": "what channel is paramount tv on direct 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: when does captain america the first avenger take place?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1942" ] }, "style": "rule" }
{ "index": 231, "need_tools_kwargs": true, "question": "when does captain america the first avenger take place?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1942" ] }, "question": "when does captain america the first avenger take 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: how many seasons are in dragon ball super?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "5" ] }, "style": "rule" }
{ "index": 232, "need_tools_kwargs": true, "question": "how many seasons are in dragon ball super?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "5" ] }, "question": "how many seasons are in dragon ball super?" } } } }
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 little girl on mrs doubtfire?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Mara Elizabeth Wilson" ] }, "style": "rule" }
{ "index": 233, "need_tools_kwargs": true, "question": "who played the little girl on mrs doubtfire?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Mara Elizabeth Wilson" ] }, "question": "who played the little girl on mrs doubtfire?" } } } }
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: based on the residues in the active site what family of proteases does pepsin belong to?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Acid proteases" ] }, "style": "rule" }
{ "index": 234, "need_tools_kwargs": true, "question": "based on the residues in the active site what family of proteases does pepsin belong to?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Acid proteases" ] }, "question": "based on the residues in the active site what family of proteases does pepsin belong to?" } } } }
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 song mama told me not to come?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Eric Burdon" ] }, "style": "rule" }
{ "index": 235, "need_tools_kwargs": true, "question": "who sings the song mama told me not to come?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Eric Burdon" ] }, "question": "who sings the song mama told me not to come?" } } } }
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 two types of information are encrypted by the https protocol (choose two.)?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "request/response load", "headers" ] }, "style": "rule" }
{ "index": 236, "need_tools_kwargs": true, "question": "what two types of information are encrypted by the https protocol (choose two.)?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "request/response load", "headers" ] }, "question": "what two types of information are encrypted by the https protocol (choose two.)?" } } } }
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 controlled the economy of the incan empire?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Inca rulers" ] }, "style": "rule" }
{ "index": 237, "need_tools_kwargs": true, "question": "who controlled the economy of the incan empire?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Inca rulers" ] }, "question": "who controlled the economy of the incan empire?" } } } }
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 song ooh child things are going to get easier?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Five Stairsteps" ] }, "style": "rule" }
{ "index": 238, "need_tools_kwargs": true, "question": "who sings the song ooh child things are going to get easier?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Five Stairsteps" ] }, "question": "who sings the song ooh child things are going to get easier?" } } } }
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 texting become available on cell phones?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1994" ] }, "style": "rule" }
{ "index": 239, "need_tools_kwargs": true, "question": "when did texting become available on cell phones?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1994" ] }, "question": "when did texting become available on cell phones?" } } } }
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 three point shot start in high school?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1987" ] }, "style": "rule" }
{ "index": 240, "need_tools_kwargs": true, "question": "when did the three point shot start in high school?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1987" ] }, "question": "when did the three point shot start in high school?" } } } }
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 commanded british forces at the battle of waterloo?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the Duke of Wellington" ] }, "style": "rule" }
{ "index": 241, "need_tools_kwargs": true, "question": "who commanded british forces at the battle of waterloo?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the Duke of Wellington" ] }, "question": "who commanded british forces at the battle of waterloo?" } } } }
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 hosted the daily show before trevor noah?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jon Stewart", "Craig Kilborn" ] }, "style": "rule" }
{ "index": 242, "need_tools_kwargs": true, "question": "who hosted the daily show before trevor noah?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jon Stewart", "Craig Kilborn" ] }, "question": "who hosted the daily show before trevor noah?" } } } }
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 elf in rudolph the red nosed reindeer?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Hermey" ] }, "style": "rule" }
{ "index": 243, "need_tools_kwargs": true, "question": "the elf in rudolph the red nosed reindeer?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Hermey" ] }, "question": "the elf in rudolph the red nosed reindeer?" } } } }
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 water in rockhampton come from?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "The Fitzroy River Barrage" ] }, "style": "rule" }
{ "index": 244, "need_tools_kwargs": true, "question": "where does the water in rockhampton come from?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "The Fitzroy River Barrage" ] }, "question": "where does the water in rockhampton 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: when did rudolph's shiny new year come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1976" ] }, "style": "rule" }
{ "index": 245, "need_tools_kwargs": true, "question": "when did rudolph's shiny new year come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1976" ] }, "question": "when did rudolph's shiny new year 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: what is considered middle age in the united states?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "45 to 65" ] }, "style": "rule" }
{ "index": 246, "need_tools_kwargs": true, "question": "what is considered middle age in the united states?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "45 to 65" ] }, "question": "what is considered middle age in the united states?" } } } }
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 liquid in a magic 8 ball?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "alcohol" ] }, "style": "rule" }
{ "index": 247, "need_tools_kwargs": true, "question": "what is the liquid in a magic 8 ball?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "alcohol" ] }, "question": "what is the liquid in a magic 8 ball?" } } } }
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 we don't have to take our clothes off original?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jermaine Stewart" ] }, "style": "rule" }
{ "index": 248, "need_tools_kwargs": true, "question": "who sings we don't have to take our clothes off original?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jermaine Stewart" ] }, "question": "who sings we don't have to take our clothes off original?" } } } }
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 gave britain the right to rule palestine?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "the League of Nations" ] }, "style": "rule" }
{ "index": 249, "need_tools_kwargs": true, "question": "who gave britain the right to rule palestine?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "the League of Nations" ] }, "question": "who gave britain the right to rule palestine?" } } } }
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 for republic of doyle?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Great Big Sea" ] }, "style": "rule" }
{ "index": 250, "need_tools_kwargs": true, "question": "who sings the theme song for republic of doyle?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Great Big Sea" ] }, "question": "who sings the theme song for republic of doyle?" } } } }
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 comprised a third of brazil's european immigrant population?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Portuguese" ] }, "style": "rule" }
{ "index": 251, "need_tools_kwargs": true, "question": "who comprised a third of brazil's european immigrant population?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Portuguese" ] }, "question": "who comprised a third of brazil's european immigrant population?" } } } }
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 blood in the pulmonary artery go?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "lungs" ] }, "style": "rule" }
{ "index": 252, "need_tools_kwargs": true, "question": "where does the blood in the pulmonary artery go?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "lungs" ] }, "question": "where does the blood in the pulmonary artery go?" } } } }
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 the movie turner and hooch what kind of dog was hooch?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Dogue de Bordeaux" ] }, "style": "rule" }
{ "index": 253, "need_tools_kwargs": true, "question": "in the movie turner and hooch what kind of dog was hooch?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Dogue de Bordeaux" ] }, "question": "in the movie turner and hooch what kind of dog was hooch?" } } } }
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 star wars movie came out before the last jedi?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "The Empire Strikes Back" ] }, "style": "rule" }
{ "index": 254, "need_tools_kwargs": true, "question": "what star wars movie came out before the last jedi?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "The Empire Strikes Back" ] }, "question": "what star wars movie came out before the last jedi?" } } } }
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 brennan get pregnant the first time?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "season six finale" ] }, "style": "rule" }
{ "index": 255, "need_tools_kwargs": true, "question": "when does brennan get pregnant the first time?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "season six finale" ] }, "question": "when does brennan get pregnant the first 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: who wrote the song drive by the cars?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Ric Ocasek" ] }, "style": "rule" }
{ "index": 256, "need_tools_kwargs": true, "question": "who wrote the song drive by the cars?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Ric Ocasek" ] }, "question": "who wrote the song drive by the cars?" } } } }
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 killer in the book ordeal by innocence?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Kirsten Lindstrom" ] }, "style": "rule" }
{ "index": 257, "need_tools_kwargs": true, "question": "who is the killer in the book ordeal by innocence?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Kirsten Lindstrom" ] }, "question": "who is the killer in the book ordeal by innocence?" } } } }
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 statue of liberty holding in her left arm?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "tabula ansata" ] }, "style": "rule" }
{ "index": 258, "need_tools_kwargs": true, "question": "what is the statue of liberty holding in her left arm?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "tabula ansata" ] }, "question": "what is the statue of liberty holding in her left 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: who plays lilly on the young and the restless?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Christel Khalil" ] }, "style": "rule" }
{ "index": 259, "need_tools_kwargs": true, "question": "who plays lilly on the young and the restless?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Christel Khalil" ] }, "question": "who plays lilly on 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: when is the new episode of suits airing?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "July 18, 2018" ] }, "style": "rule" }
{ "index": 260, "need_tools_kwargs": true, "question": "when is the new episode of suits airing?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "July 18, 2018" ] }, "question": "when is the new episode of suits airing?" } } } }
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 costco's are there in the united states?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "519" ] }, "style": "rule" }
{ "index": 261, "need_tools_kwargs": true, "question": "how many costco's are there in the united states?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "519" ] }, "question": "how many costco's are there in the united states?" } } } }
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 electron transport chain of photosynthesis located?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "sunlight" ] }, "style": "rule" }
{ "index": 262, "need_tools_kwargs": true, "question": "where is the electron transport chain of photosynthesis located?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "sunlight" ] }, "question": "where is the electron transport chain of photosynthesis located?" } } } }
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 full game fortnite come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "in 2018" ] }, "style": "rule" }
{ "index": 263, "need_tools_kwargs": true, "question": "when does the full game fortnite come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "in 2018" ] }, "question": "when does the full game fortnite 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: how many us presidents have there been assassinated?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Four" ] }, "style": "rule" }
{ "index": 264, "need_tools_kwargs": true, "question": "how many us presidents have there been assassinated?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Four" ] }, "question": "how many us presidents have there been assassinated?" } } } }
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 actor that plays lucius malfoy?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jason Isaacs" ] }, "style": "rule" }
{ "index": 265, "need_tools_kwargs": true, "question": "who is the actor that plays lucius malfoy?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jason Isaacs" ] }, "question": "who is the actor that plays lucius malfoy?" } } } }
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 will on girlfriends guide to divorce?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Warren Christie" ] }, "style": "rule" }
{ "index": 266, "need_tools_kwargs": true, "question": "who played will on girlfriends guide to divorce?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Warren Christie" ] }, "question": "who played will on girlfriends guide to divorce?" } } } }
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 house in the movie belly?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Jamaica Estates" ] }, "style": "rule" }
{ "index": 267, "need_tools_kwargs": true, "question": "where is the house in the movie belly?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Jamaica Estates" ] }, "question": "where is the house in the movie belly?" } } } }
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 telephone area code 585 located?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Rochester" ] }, "style": "rule" }
{ "index": 268, "need_tools_kwargs": true, "question": "where is the telephone area code 585 located?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Rochester" ] }, "question": "where is the telephone area code 585 located?" } } } }
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 earl's dad in my name is earl?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Beau Bridges" ] }, "style": "rule" }
{ "index": 269, "need_tools_kwargs": true, "question": "who plays earl's dad in my name is earl?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Beau Bridges" ] }, "question": "who plays earl's dad in my name is earl?" } } } }
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 movie dirty dancing come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "August 21, 1987" ] }, "style": "rule" }
{ "index": 270, "need_tools_kwargs": true, "question": "when did the movie dirty dancing come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "August 21, 1987" ] }, "question": "when did the movie dirty dancing 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 is the first lady of the usa?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Melania Trump" ] }, "style": "rule" }
{ "index": 271, "need_tools_kwargs": true, "question": "who is the first lady of the usa?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Melania Trump" ] }, "question": "who is the first lady of the usa?" } } } }
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 amys brother on everybody loves raymond?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Peter MacDougall" ] }, "style": "rule" }
{ "index": 272, "need_tools_kwargs": true, "question": "who plays amys brother on everybody loves raymond?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Peter MacDougall" ] }, "question": "who plays amys brother on everybody 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: who sang come and get your love now?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Redbone" ] }, "style": "rule" }
{ "index": 273, "need_tools_kwargs": true, "question": "who sang come and get your love now?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Redbone" ] }, "question": "who sang come and get your love now?" } } } }
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 king james bible translated to english?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1611" ] }, "style": "rule" }
{ "index": 274, "need_tools_kwargs": true, "question": "when was the king james bible translated to english?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1611" ] }, "question": "when was the king james bible translated to english?" } } } }
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 last name dominguez come from?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "a name of Spanish origin" ] }, "style": "rule" }
{ "index": 275, "need_tools_kwargs": true, "question": "where did the last name dominguez come from?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "a name of Spanish origin" ] }, "question": "where did the last name dominguez 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: how many super bowl wins does elway have?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "three" ] }, "style": "rule" }
{ "index": 276, "need_tools_kwargs": true, "question": "how many super bowl wins does elway have?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "three" ] }, "question": "how many super bowl wins does elway have?" } } } }
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 leprosy most commonly found in the world?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "India" ] }, "style": "rule" }
{ "index": 277, "need_tools_kwargs": true, "question": "where is leprosy most commonly found in the world?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "India" ] }, "question": "where is leprosy most commonly found in the world?" } } } }
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 crowned himself as an emperor of france in 1804?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Napoleon Bonaparte" ] }, "style": "rule" }
{ "index": 278, "need_tools_kwargs": true, "question": "who crowned himself as an emperor of france in 1804?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Napoleon Bonaparte" ] }, "question": "who crowned himself as an emperor of france in 1804?" } } } }
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 song the ballad of lucy jordan?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Shel Silverstein" ] }, "style": "rule" }
{ "index": 279, "need_tools_kwargs": true, "question": "who wrote the song the ballad of lucy jordan?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Shel Silverstein" ] }, "question": "who wrote the song the ballad of lucy jordan?" } } } }
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: japanese baseball players in mlb hall of fame?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Hideo Nomo", "Kazuhiro Sasaki", "Hideki Matsui" ] }, "style": "rule" }
{ "index": 280, "need_tools_kwargs": true, "question": "japanese baseball players in mlb hall of fame?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Hideo Nomo", "Kazuhiro Sasaki", "Hideki Matsui" ] }, "question": "japanese baseball players in mlb hall of fame?" } } } }
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: greek goddess of peace with the roman equivalent pax?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Eirene" ] }, "style": "rule" }
{ "index": 281, "need_tools_kwargs": true, "question": "greek goddess of peace with the roman equivalent pax?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Eirene" ] }, "question": "greek goddess of peace with the roman equivalent pax?" } } } }
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 original singer of one day?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Matisyahu" ] }, "style": "rule" }
{ "index": 282, "need_tools_kwargs": true, "question": "who is the original singer of one day?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Matisyahu" ] }, "question": "who is the original singer of one day?" } } } }
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 jason put on the hockey mask?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1980" ] }, "style": "rule" }
{ "index": 283, "need_tools_kwargs": true, "question": "when did jason put on the hockey mask?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1980" ] }, "question": "when did jason put on the hockey mask?" } } } }
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 gone with the wind won 10 academy awards?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "At the 12th Academy Awards" ] }, "style": "rule" }
{ "index": 284, "need_tools_kwargs": true, "question": "when did gone with the wind won 10 academy awards?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "At the 12th Academy Awards" ] }, "question": "when did gone with the wind won 10 academy awards?" } } } }
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 cartman's voice on south park?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Trey Parker" ] }, "style": "rule" }
{ "index": 285, "need_tools_kwargs": true, "question": "who does cartman's voice on south park?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Trey Parker" ] }, "question": "who does cartman's voice on south park?" } } } }
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 carol end up with on er?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Doug" ] }, "style": "rule" }
{ "index": 286, "need_tools_kwargs": true, "question": "who does carol end up with on er?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Doug" ] }, "question": "who does carol end up with on er?" } } } }
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 long distance phone call made?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "between Cambridge and Boston" ] }, "style": "rule" }
{ "index": 287, "need_tools_kwargs": true, "question": "where was the first long distance phone call made?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "between Cambridge and Boston" ] }, "question": "where was the first long distance phone call made?" } } } }
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 are the traditional appetizers in spain called?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Tapas" ] }, "style": "rule" }
{ "index": 288, "need_tools_kwargs": true, "question": "what are the traditional appetizers in spain called?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Tapas" ] }, "question": "what are the traditional appetizers in spain called?" } } } }
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 current inspector general of police of uganda?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Martin Okoth Ochola" ] }, "style": "rule" }
{ "index": 289, "need_tools_kwargs": true, "question": "who is the current inspector general of police of uganda?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Martin Okoth Ochola" ] }, "question": "who is the current inspector general of police of uganda?" } } } }
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 in lucy in the sky with diamonds?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Paul McCartney – harmony vocals" ] }, "style": "rule" }
{ "index": 290, "need_tools_kwargs": true, "question": "who sings in lucy in the sky with diamonds?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Paul McCartney – harmony vocals" ] }, "question": "who sings in lucy in the sky with diamonds?" } } } }
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 vikings made the playoffs?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "2017" ] }, "style": "rule" }
{ "index": 291, "need_tools_kwargs": true, "question": "when was the last time vikings made the playoffs?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "2017" ] }, "question": "when was the last time vikings made 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 does disney's live action aladdin come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "May 24, 2019" ] }, "style": "rule" }
{ "index": 292, "need_tools_kwargs": true, "question": "when does disney's live action aladdin come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "May 24, 2019" ] }, "question": "when does disney's live action aladdin 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: when did the song body like a backroad come out?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "February 1, 2017" ] }, "style": "rule" }
{ "index": 293, "need_tools_kwargs": true, "question": "when did the song body like a backroad come out?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "February 1, 2017" ] }, "question": "when did the song body like a backroad 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: muzon san jose del monte bulacan zip code?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "3023, 3024" ] }, "style": "rule" }
{ "index": 294, "need_tools_kwargs": true, "question": "muzon san jose del monte bulacan zip code?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "3023, 3024" ] }, "question": "muzon san jose del monte bulacan zip code?" } } } }
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 next episode of superstore air?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "on October 4, 2018" ] }, "style": "rule" }
{ "index": 295, "need_tools_kwargs": true, "question": "when does the next episode of superstore air?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "on October 4, 2018" ] }, "question": "when does the next episode of superstore air?" } } } }
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 american society of anesthesiologists founded?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "1936" ] }, "style": "rule" }
{ "index": 296, "need_tools_kwargs": true, "question": "when was the american society of anesthesiologists founded?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "1936" ] }, "question": "when was the american society of anesthesiologists founded?" } } } }
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 are the two major regions where dense water masses are created?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "North Atlantic", "Southern Ocean" ] }, "style": "rule" }
{ "index": 297, "need_tools_kwargs": true, "question": "where are the two major regions where dense water masses are created?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "North Atlantic", "Southern Ocean" ] }, "question": "where are the two major regions where dense water masses are created?" } } } }
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 team does steph curry brother play for?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Dallas Mavericks" ] }, "style": "rule" }
{ "index": 298, "need_tools_kwargs": true, "question": "what team does steph curry brother play for?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Dallas Mavericks" ] }, "question": "what team does steph curry brother play for?" } } } }
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 closest airport to white sulphur springs west virginia?", "role": "user" } ]
fact-reasoning
{ "ground_truth": { "target": [ "Greenbrier Valley Airport" ] }, "style": "rule" }
{ "index": 299, "need_tools_kwargs": true, "question": "what is the closest airport to white sulphur springs west virginia?", "split": "test", "tools_kwargs": { "search": { "create_kwargs": { "data_source": "nq", "ground_truth": { "target": [ "Greenbrier Valley Airport" ] }, "question": "what is the closest airport to white sulphur springs west virginia?" } } } }
null