id string | slice string | category string | enable_thinking bool | messages list | tools string | n_tokens int64 | n_loss_tokens int64 | epochs int64 | source_dataset string | source_id string | source_revision string | source_url string | source_title string | followup_source string | prompt_licence string | problem_author string | answer_generator string | seed_generator string | row_licence string | ai_generated bool | created_with_grok bool | label string | verification string | text_omitted bool | text_omitted_reason string | pii_redactions list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
chat-rp-0d9f618a47c5 | chat | rewriting_editing | false | [
{
"role": "user",
"content": "rewrite: In recent studies, in order to improve the cell culture media, platelet-rich plasma (PRP) has been used as a supplement, which has been shown to enhance the proliferation, migration, and differentiation of stem cells and help the tissue regeneration process.\nPlatelets... | null | 173 | 81 | 1 | lmarena-ai/arena-human-preference-100k | 7a9324424cdf4f01933554e927829605 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
chat-rp-642ef6aeb554 | chat | brainstorming_lists | false | [
{
"role": "user",
"content": "Give me 15 name ideas for account on scratch where i would do test projects.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Here are 15 Scratch username ideas for a test-proje... | null | 195 | 168 | 1 | allenai/WildChat-4.8M | 60b70be3d5e88e962d97ee9fa27f3db5:1938266 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
cthink-rp-48a7512444ce | chat_thinking | explain_teach | true | [
{
"role": "user",
"content": "Why do people speedcube?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "People speedcube for a mix of challenge, progress and community:\n\n- **The skill ceiling.** Solving a ... | null | 336 | 312 | 1 | lmarena-ai/arena-human-preference-100k | ffe2f2a33b6b46b4bc4d95f781681032 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | length + linear-thinking rules; LLM judge: score>=4, answer follows thinking, thinking clean | false | null | null |
sysgen-0038-0-0 | system_prompts | generated_in_scope | false | [
{
"role": "system",
"content": "You are the virtual counter hand for Ember & Oak Coffee Roasters, a small-batch roastery in Leeds. We are open Tuesday to Saturday, 8am to 4pm, and a 250g bag of our House Blend costs £9.50 while single-origin bags are £12. We do not ship internationally. Keep replies warm, b... | null | 131 | 24 | 1 | generated by DeepSeek-V4.1-Flash | sysgen-0038-0-0 | null | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (prompt and answers) | LLM quality screen (score>=4) | false | null | null |
tool-gsm8k-train-390 | python_tool | math | false | [
{
"role": "user",
"content": "Oscar wants to train for a marathon. He plans to add 2/3 of a mile each week until he reaches a 20-mile run. How many weeks before the marathon should he start training if he has already run 2 miles?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": n... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 189 | 60 | 1 | openai/gsm8k | gsm8k-train-390 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
chat-rp-279cf62ef832 | chat | practical_advice | false | [
{
"role": "user",
"content": "Interview rating scales explained 1-5",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "A 1–5 rating scale usually works like this (anchors vary by employer, so check the rubric)... | null | 286 | 268 | 1 | allenai/WildChat-4.8M | a2a1029ff621e1af186ead0d3844d96b:1515841 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
gen-mt-nemo-chat-001982 | gen_mt | shorter | false | [
{
"role": "user",
"content": "when running between wickets in cricket, can you run on the pitch or do you have to leave it immediately?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "In cricket, batsmen **... | null | 454 | 409 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-001982 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
if-rp-bfde7c58b0ee | instruction_following | writing_create | false | [
{
"role": "user",
"content": "Please help me compose email to interviewees I would like to interview for weekend reception role",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Subject: Interview invitation ... | null | 465 | 404 | 1 | allenai/WildChat-4.8M | b4080df1e446afb27b5569fcb829f731:1948140 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); follow-up user turn templated by Udgam Labs code | programmatic constraint checkers (all must pass) + LLM quality screen (score>=4) | false | null | null |
chat-rp-e888c5c7936b | chat | code_help | false | [
{
"role": "user",
"content": "how to train RT-DETRv2 on custom dataset in python",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "RT-DETRv2 isn't in Ultralytics yet, so the realistic path is the official Pa... | null | 791 | 735 | 1 | allenai/WildChat-4.8M | 65a7aee55227186f91671f5c5681a554:4037781 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | generated | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
if-rp-b7ee8e3d570b | instruction_following | explain_teach | false | [
{
"role": "user",
"content": "Write your entire reply in capital letters. Format the answer as exactly 5 bullet points, each starting with \"- \". Use the words \"simple\" and \"quickly\" somewhere in your reply.\n\nWhat does hack the simulation mean?",
"reasoning_content": null,
"tool_calls": null,... | null | 471 | 415 | 1 | lmarena-ai/arena-human-preference-100k | 6c0d02b4745a48269e4e44b7b2b468b4 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | programmatic constraint checkers (all must pass) + LLM quality screen (score>=4) | false | null | null |
math-gsm8k-train-2264-think-nothink | maths_logic | gsm8k_train | false | [
{
"role": "user",
"content": "Sun City has 1000 more than twice as many people as Roseville City. Roseville city has 500 less than thrice as many people as Willowdale city. If Willowdale city has 2000 people, how many people live in Sun City?",
"reasoning_content": null,
"tool_calls": null,
"too... | null | 120 | 52 | 1 | openai/gsm8k | gsm8k-train-2264 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
gen-mt-nemo-chat-001022 | gen_mt | another | false | [
{
"role": "user",
"content": "can try to generate a complete, newly created tune in ABC notation with a full structure,\nIn the tune",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Here's a newly created tu... | null | 690 | 643 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-001022 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
gen-mt-nemo-chat-025291 | gen_mt | switch | false | [
{
"role": "user",
"content": "relate ds descroption with heading -MERN Stack Development\nDeliver cross-platform and responsive applications, enabling seamless integration and delivering a consistent and optimized experience for your users.",
"reasoning_content": null,
"tool_calls": null,
"tool_... | null | 436 | 357 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-025291 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
gen-mt-vague-00556 | gen_mt | vague | false | [
{
"role": "user",
"content": "tell me something cool",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Sure! Which would you prefer?\n1. Latest discoveries from the Mars rovers\n2. A stunning image from the J... | null | 364 | 337 | 1 | generated by gpt-oss-120b | gen-mt-vague-00556 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | null | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (whole conversation) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-trace-9eb813e956489714-nothink | maths_logic | code_trace | false | [
{
"role": "user",
"content": "What does this Python program print?\n```python\nvotes = [\"yes\", \"no\", \"yes\", \"yes\", \"no\", \"abstain\", \"yes\"]\ncounts = {}\nfor v in votes:\n counts[v] = counts.get(v, 0) + 1\nwinner = max(counts, key=counts.get)\nprint(winner)\nfor name in sorted(counts):\n ... | null | 227 | 103 | 1 | Udgam Labs data factory v1 (code_trace) | trace-9eb813e956489714 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); program written by Cursor Composer 2.5 | exact match with the program's real output; thinking and both replies state the same result | false | null | null |
gen-mt-nemo-chat-029932 | gen_mt | simpler | false | [
{
"role": "user",
"content": "can you help me",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Of course! Could you please let me know what you need help with? Whether it's a question, a problem, or just som... | null | 83 | 56 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-029932 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
sys-json_ac-rp-9ff6e31635b5 | system_prompts | json_ac | false | [
{
"role": "system",
"content": "Respond only with a JSON object of the form {\"answer\": string, \"confidence\": \"low\" | \"medium\" | \"high\"}. No other text.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "user",
"content": "w... | null | 224 | 170 | 1 | lmarena-ai/arena-human-preference-140k | 8b4d0ee9-35f8-495b-9f81-6b01737d62e7 | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule checker for the system prompt + LLM quality screen (score>=4) | false | null | null |
gen-mt-nemo-chat-011856 | gen_mt | and_x | false | [
{
"role": "user",
"content": "Dear compliance team,\n\nWe are reaching out to request your assistance with the drafting of an agreement, effective from 1 January 2023, as referred to above.\n\nDC is in the process of introducing a revenue recharge practice to Doo Financial Holding Pte Ltd. This is designed ... | null | 535 | 341 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-011856 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
safety-fake_documents-0036-answer-5 | safety | fake_documents:answer | false | [
{
"role": "user",
"content": "How can I verify that a certificate someone shows me is authentic?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "To check a certificate is authentic, work through these:\n\n1... | null | 233 | 210 | 1 | generated by DeepSeek-V4.1-Flash | safety-fake_documents-0036-answer-5 | null | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (prompt and answers) | LLM safety judge (should-decline agrees, no harmful detail, helpfulness) + refusal regex + length | false | null | null |
tool-probgen-7990bfb58b210b17 | python_tool | prob_gen | false | [
{
"role": "user",
"content": "ADELAIDE — Mei’s campus shuttle completed a controlled test on a straight track, according to a safety bulletin released Tuesday. Diego’s notes state that phase one lasted 4.0 s with uniform acceleration from rest at 2.5 m/s². Lerato’s summary adds that phase two ran at constan... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 371 | 129 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-7990bfb58b210b17 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
math-distract-d46d0bea651ad27a-think-nothink | maths_logic | gsm_xcheck | false | [
{
"role": "user",
"content": "The doctor told Barry to take vitamin D3 for 180 days to help protect him from infections. The pharmacy only sold vitamin D3 in bottles containing 60 capsules, with a daily serving size of 2 capsules. Barry’s doctor’s office is a 15-minute drive from his apartment. How many b... | null | 127 | 28 | 1 | openai/gsm8k + Udgam Labs data factory (distract) | gsm8k-train-3384 | distract-d46d0bea651ad27a | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT AND CC-BY-4.0 | openai/gsm8k (train) + one sentence by cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); GSM8K train problem with one added sentence written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
chat-rp-61b9c9fbc5e6 | chat | code_help | false | [
{
"role": "user",
"content": "does the floyd-warshall algorithm give the correct result if the input graph has negative weights (but no negative cycle)? breieflly explain your answer.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "as... | null | 720 | 616 | 1 | lmarena-ai/arena-human-preference-100k | cac16013891a4e7d93ea6b87ca9360bd | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | generated | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
chat-rp-bffa100bf78b | chat | practical_advice | false | [
{
"role": "user",
"content": "how to write an email to professor about my career concern between styling and fashion design",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Write a short, specific email that... | null | 343 | 316 | 1 | allenai/WildChat-4.8M | 21a68eb82ffea0ab996eba3c2b6566b7:1504006 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
chat-rp-9f4a917593c9 | chat | brainstorming_lists | false | [
{
"role": "user",
"content": "give me a list of the consonants and vowels that italian speakers struggle with when speaking English along with their phonetic symbols and a brief explanation why.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
... | null | 411 | 369 | 1 | allenai/WildChat-4.8M | 46c987941a0fa49c78aa75f9a3644efb:2990313 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
tool-gsm8k-train-1738 | python_tool | math | false | [
{
"role": "user",
"content": "Steve has decided to become a baker. On Mondays, Wednesdays and Fridays, he bakes apple pies. On Tuesdays and Thursdays, he bakes cherry pies. If he bakes 12 pies per day, how many more apple pies does he bake than cherry pies in one week?",
"reasoning_content": null,
"... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 251 | 99 | 1 | openai/gsm8k | gsm8k-train-1738 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
safety-gambling_scams-0040-decline-0 | safety | gambling_scams:decline | false | [
{
"role": "user",
"content": "My teen found a way to rig the bets on a gambling app and wants to try it, how do I stop him?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "I can't help with rigging or tampe... | null | 89 | 54 | 1 | generated by DeepSeek-V4.1-Flash | safety-gambling_scams-0040-decline-0 | null | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (prompt and answers) | LLM safety judge (should-decline agrees, no harmful detail, helpfulness) + refusal regex + length | false | null | null |
cthink-rp-abdddccbb02e | chat_thinking | brainstorming_lists | true | [
{
"role": "user",
"content": "I can turn one object into gold. what object is the best choice? Just name your suggestion.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "A large rock — ideally a boulder you... | null | 228 | 191 | 1 | lmarena-ai/arena-human-preference-100k | f2594e7381ed44a58fb582a2a3995cb5 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | length + linear-thinking rules; LLM judge: score>=4, answer follows thinking, thinking clean | false | null | null |
tool-probgen-192cdd49baf44fed | python_tool | gsm_xcheck | false | [
{
"role": "user",
"content": "Lars, Hao, and Keoni each have one published score on a charity leaderboard, measured in points: 8,400; 7,100; and 9,500 respectively. Before publication, two earlier entries by the same athletes were removed. Those two removed entries were equal and were the lowest scores in t... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 359 | 164 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-192cdd49baf44fed | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
math-probgen-0b0a143bb9d4d509-think | maths_logic | prob_gen | true | [
{
"role": "user",
"content": "Solve: LAB NOTES — Takoda, Lucas, and Wanjiru prepared saline for a field test. Only a stock jug of 35% saline and pure water (0% salt) were used. They poured all of the stock into a beaker, added 540 mL of water, and obtained exactly 900 mL of mixed solution with no spillage. ... | null | 249 | 126 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-0b0a143bb9d4d509 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
gen-mt-synth-009670 | gen_mt | switch | false | [
{
"role": "user",
"content": "The historical use of wood gas technology for locomotive fuel was most prominently observed in which of the following regions?\n\nA. In France and Italy, where it was a notable but not exclusive practice.\nB. In Great Britain, as a primary and widespread alternative to coal.\nC... | null | 160 | 26 | 1 | PleIAs/SYNTH | synth-009670 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-mcq | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-probgen-d4b01d42b96da4ae-think-nothink | maths_logic | prob_gen | false | [
{
"role": "user",
"content": "Solve: Steam curled off the teacups at a Chengdu lunch. Elena, 20 years old this month, listened as Anna described a school fair. Anna is 75% of Elena's age, and the fair begins in a number of years equal to 25% of Elena's present age. Noa, who is 40% of Elena's age, folded a p... | null | 151 | 29 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-d4b01d42b96da4ae | null | null | null | null | CC-BY-4.0 | xai/grok-4.7 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | true | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by xAI Grok 4.7. Created with Grok | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
gen-mt-synth-000741 | gen_mt | dig | false | [
{
"role": "user",
"content": "How would the cumulative effect of NHEJ-mediated mutations throughout an individual's lifetime likely impact their genetic stability and risk of developing hereditary disorders?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
... | null | 396 | 344 | 1 | PleIAs/SYNTH | synth-000741 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-memorization | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-gsm8k-train-2615-think-nothink | maths_logic | gsm8k_train | false | [
{
"role": "user",
"content": "Six bottles of 2 liters of water cost $12. What is the price of 1 liter of water? Please show your working.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "The six bottles hold... | null | 68 | 31 | 1 | openai/gsm8k | gsm8k-train-2615 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
sys-kid7-rp-7641cfd96026 | system_prompts | kid7 | false | [
{
"role": "system",
"content": "You explain things to a curious seven-year-old: short sentences, everyday words, one fun comparison.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "user",
"content": "why does tuberculosis makes a ... | null | 155 | 109 | 1 | lmarena-ai/arena-human-preference-100k | 56bb1bd73f73454c8a1745d1efd76cfc | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule checker for the system prompt + LLM quality screen (score>=4) | false | null | null |
long-54d60018c1-q0 | long_documents | wikipedia | false | [
{
"role": "user",
"content": "What did President Biden announce on 2 May 2022 regarding batteries?\n\n# Electric vehicle battery\n\nAn electric vehicle battery (EVB, also known as a traction battery) is a rechargeable battery used to power the electric motors of a battery electric vehicle (BEV) or hybrid el... | null | 10,538 | 66 | 1 | wikimedia/wikipedia (20231101.en) | wikipedia:15452708 | b04c8d1ceb2f5cd4588862100d08de323dccfbaa | https://en.wikipedia.org/wiki/Electric%20vehicle%20battery | Electric vehicle battery | null | CC-BY-SA-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-SA-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); text from Wikipedia (CC-BY-SA-4.0, see source_title and source_url) | answer grounded: supporting quote found verbatim in the document | false | null | null |
chat-rp-4ccc0c8e90ba | chat | practical_advice | false | [
{
"role": "user",
"content": "Where should I put a state bar license on my resume?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Put your state bar license in a **\"Bar Admissions\" section**, placed near... | null | 198 | 176 | 1 | lmarena-ai/arena-human-preference-100k | 96db838abe86483ba320477c69775a23 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
fc-01127 | json_function_calls | parallel | false | [
{
"role": "user",
"content": "How much would 750 euros get me in Japanese yen and in Swiss francs right now?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "",
"reasoning_content": null,
"tool_calls... | [{"type": "function", "function": {"name": "rate_lookup", "description": "Fetches the current mid-market exchange rate for a currency pair.", "parameters": {"type": "object", "properties": {"base_currency": {"type": "string", "description": "Three-letter ISO code of the currency being sold."}, "quote_currency": {"type"... | 380 | 103 | 1 | generated by DeepSeek-V4.1-Flash | fc-01127 | null | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (prompt and answers) | tool schemas validated; every call validated against its schema and mock-executed (scripted result); final reply uses the results | false | null | null |
math-probgen-da8994c525d5b266-nothink | maths_logic | prob_gen | false | [
{
"role": "user",
"content": "Hi — I’m Dorji in Cape Town and I’m stuck on our shop invoice (all amounts in rand). Kamal and Noah split things with me. The main display was R 84,000 before 15% VAT; Kamal was supposed to cover 40% of the VAT-inclusive price of that display only. For cabling, the quote was R ... | null | 351 | 182 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-da8994c525d5b266 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
if-rp-e285e0f3a4ad | instruction_following | brainstorming_lists | false | [
{
"role": "user",
"content": "What are some of the pros and cons of Lencioni’s Ideal Team Player Model of Humble as it relates to recruiting high performing team members",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
... | null | 998 | 934 | 1 | allenai/WildChat-4.8M | c9bd79e6274daef181b30f038808abe5:1580187 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); follow-up user turn templated by Udgam Labs code | programmatic constraint checkers (all must pass) + LLM quality screen (score>=4) | false | null | null |
chat-rp-36d92ababd41 | chat | brainstorming_lists | false | [
{
"role": "user",
"content": "What does the following mean in simplified terms:I recommend we take the position that the provincial legislation is declaratory of the common law.\nThis allows us to express that s71 of the Legislation Act is the starting point for the analysis, allowing us to ground a positio... | null | 370 | 187 | 1 | lmarena-ai/arena-human-preference-100k | cae8b20c3d184416b837906243f93cfe | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
chat-rp-11e113f52096 | chat | practical_advice | false | [
{
"role": "user",
"content": "how to learn Executive presence in detailed poin with example for each? Along with how to be more precise in my replying ?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "**Exe... | null | 406 | 370 | 1 | allenai/WildChat-4.8M | fcfa8436388fba764ef0af8e46e4c8cd:3989997 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
gen-mt-synth-002441 | gen_mt | refine | false | [
{
"role": "user",
"content": "How come these cookies keep ending up on my computer when I visit websites - like, why do I get them and what do they do?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Cookie... | null | 369 | 298 | 1 | PleIAs/SYNTH | synth-002441 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-memorization | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
chat-rp-7d1acaf40ed0 | chat | brainstorming_lists | false | [
{
"role": "user",
"content": "Come up with ideas for a creative awareness campaign about the fact that hypertension has no symptoms.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "1. **\"The Quiet Room\" p... | null | 562 | 509 | 1 | lmarena-ai/arena-human-preference-100k | 92de9cd2be2c476589e6339b546fb843 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | generated | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
math-gsm8k-train-3011-think | maths_logic | gsm8k_train | true | [
{
"role": "user",
"content": "Thomas is training at the gym to prepare for a competition. He trained for 5 hours every day for a month (30 days). If he continues to train for the next 12 days, how many hours will he spend on training in total?",
"reasoning_content": null,
"tool_calls": null,
"to... | null | 188 | 121 | 1 | openai/gsm8k | gsm8k-train-3011 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
chat-rp-67fdcde8dc29 | chat | code_help | false | [
{
"role": "user",
"content": "how t o run django server",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Run from the folder containing `manage.py`:\n\n```bash\npython manage.py runserver\n```\n\nThe dev ser... | null | 160 | 144 | 1 | allenai/WildChat-4.8M | 6191472e46126e5ff2e4c4a3bdf1707a:564865 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
long-f1af8445bb-sum | long_documents | wikipedia | false | [
{
"role": "user",
"content": "Summarise this document in about 150 words.\n\n# Sabine Lisicki\n\nSabine Katharina Lisicki (; born 22 September 1989) is a German professional tennis player.\n\nLisicki turned professional in 2006, and her breakthrough came in 2009 when she reached the quarterfinals of the Wim... | null | 14,070 | 258 | 1 | wikimedia/wikipedia (20231101.en) | wikipedia:15230658 | b04c8d1ceb2f5cd4588862100d08de323dccfbaa | https://en.wikipedia.org/wiki/Sabine%20Lisicki | Sabine Lisicki | null | CC-BY-SA-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-SA-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); text from Wikipedia (CC-BY-SA-4.0, see source_title and source_url); follow-up user turns by DeepSeek-V4.1-Flash | summary length checked; answers grounded by a verbatim quote from the document | false | null | null |
chat-rp-bb7b7777e5c8 | chat | factual_qa | false | [
{
"role": "user",
"content": "what is \"bootstrapping\"?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Bootstrapping means starting and growing something using only your own resources, without outside hel... | null | 426 | 380 | 1 | lmarena-ai/arena-human-preference-140k | 3c6598dd-fb58-43b3-a6dc-42511de34d50 | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | generated | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
math-gsm8k-train-6811-think | maths_logic | gsm8k_train | true | [
{
"role": "user",
"content": "There are 24 marbles in a jar. Half are blue. There are 6 red marbles. The rest of the marbles are orange. How many orange marbles are there?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
... | null | 146 | 95 | 1 | openai/gsm8k | gsm8k-train-6811 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
chat-rp-d9ea5f0fc220 | chat | practical_advice | false | [
{
"role": "user",
"content": "how to differentiate :Osteomyelitis\nInfections of prosthetic implants\nDiskitis\nInfectious arthritis\nMyositis",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Differentiating... | null | 746 | 688 | 1 | allenai/WildChat-4.8M | 986f076cac0be0b5ba5c72e42ba477c3:2994732 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | real user | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); second user turn by the same real user | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
long-23324b3116-q1 | long_documents | wikipedia | false | [
{
"role": "user",
"content": "# American Century Championship\n\nThe American Century Championship (currently known as the American Century Championship for sponsorship reasons; previously known as the Celebrity Golf Tournament, the Isuzu Celebrity Golf Championship, and the American Century Celebrity Golf ... | null | 9,801 | 102 | 1 | wikimedia/wikipedia (20231101.en) | wikipedia:16143011 | b04c8d1ceb2f5cd4588862100d08de323dccfbaa | https://en.wikipedia.org/wiki/American%20Century%20Championship | American Century Championship | null | CC-BY-SA-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-SA-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); text from Wikipedia (CC-BY-SA-4.0, see source_title and source_url) | answer grounded: supporting quote found verbatim in the document | false | null | null |
chat-rp-510648569326 | chat | practical_advice | false | [
{
"role": "user",
"content": "How to add firebase analytics google in development build expo?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Add Firebase Analytics to an Expo development build using the Re... | null | 420 | 397 | 1 | allenai/WildChat-4.8M | a312f485929cf73e81bab61599d04787:4043070 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
if-rp-e6bea97f98f8 | instruction_following | writing_create | false | [
{
"role": "user",
"content": "Help me write an self introduction for my interview of the course of integrated BBA\n\nHighlight exactly 3 key phrases in **bold**. Keep your answer under 60 words.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
... | null | 92 | 48 | 1 | allenai/WildChat-4.8M | 57a507e3af0e823657496b9fa11420e4:1036769 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | programmatic constraint checkers (all must pass) + LLM quality screen (score>=4) | false | null | null |
gen-mt-nemo-chat-006077 | gen_mt | shorter | false | [
{
"role": "user",
"content": "Rephrase very thoroughly without omitting anything \"The reason is illegal wiretapping and surveillance.\nInitially, Mr. Andreev was detained for 24 hours. Subsequently, Christian\nAndreev was not charged in the context of the criminal proceedings in which he was\nauthorized se... | null | 595 | 325 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-006077 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
fc-00352 | json_function_calls | single | true | [
{
"role": "user",
"content": "What's the German translation for 'cloud computing' in the TechGlossary?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "",
"reasoning_content": "The user wants a translati... | [{"type": "function", "function": {"name": "lookup_term", "description": "Look up the translation of a term in a given glossary and language pair.", "parameters": {"type": "object", "properties": {"term": {"type": "string", "description": "The source term to look up."}, "glossary": {"type": "string", "description": "Th... | 589 | 96 | 1 | generated by DeepSeek-V4.1-Flash | fc-00352 | null | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (prompt and answers) | tool schemas validated; every call validated against its schema and mock-executed (scripted result); final reply uses the results | false | null | null |
chat-rp-f2212f9e5ef0 | chat | factual_qa | false | [
{
"role": "user",
"content": "what is sea means.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "\"Sea\" usually means a large body of salt water — smaller than an ocean, often partly enclosed by land (like... | null | 411 | 338 | 1 | allenai/WildChat-4.8M | 7ea060062788b100decaa245a5556171:1940791 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | generated | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
gen-mt-synth-018598 | gen_mt | another | false | [
{
"role": "user",
"content": "I was wondering if you could maybe put this information about the company into bullet points or something easier to read. Also it would be helpful if you could tell me what kind of text this is - like is it from a textbook or what. Here is the text: L'Oréal S.A. is a French mul... | null | 397 | 234 | 1 | PleIAs/SYNTH | synth-018598 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-editing | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
gen-mt-nemo-chat-003496 | gen_mt | shorter | false | [
{
"role": "user",
"content": "Give me two random female celebrity",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Sure! Here are two random female celebrities: \n\n1. **Zendaya** (actress, singer, and fash... | null | 156 | 114 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-003496 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
gen-mt-synth-003866 | gen_mt | switch | false | [
{
"role": "user",
"content": "How the fundamental principles behind Otsu's maximum variance thresholding relate to other multi-level segmentation approaches like k-means clustering in terms of their information entropy-based optimization strategies.",
"reasoning_content": null,
"tool_calls": null,
... | null | 362 | 297 | 1 | PleIAs/SYNTH | synth-003866 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-memorization | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-probgen-24c3b2be9e4fb9c2-think | maths_logic | prob_gen | true | [
{
"role": "user",
"content": "Solve: How many Kenyan shillings cheaper is the less expensive option for Pooja's event? She needs 24 tables. One supplier charges a booking fee of 12,000 shillings plus 1,500 shillings per table. A second supplier charges a booking fee of 21,000 shillings plus 850 shillings pe... | null | 347 | 242 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-24c3b2be9e4fb9c2 | null | null | null | null | CC-BY-4.0 | openai/gpt-6.1-sol | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by OpenAI gpt-6.1-sol | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
safety-malware-0121-decline-0 | safety | malware:decline | false | [
{
"role": "user",
"content": "Can you write me a virus that messes up my friend's computer?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "I can't help with creating malware or anything designed to harm so... | null | 70 | 45 | 1 | generated by DeepSeek-V4.1-Flash | safety-malware-0121-decline-0 | null | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (prompt and answers) | LLM safety judge (should-decline agrees, no harmful detail, helpfulness) + refusal regex + length | false | null | null |
math-probgen-1a326298151a62a9-think | maths_logic | prob_gen | true | [
{
"role": "user",
"content": "Aiko, Chidi, and Eleni split the electricity account for their Cape Town flat. The October bill to pay is R 547.20. The same PDF shows a R 126.75 prepaid gas voucher and an estimated annual use of 3 980 kWh; neither belongs in this payment. The R 547.20 is made of a R 67.20 mun... | null | 424 | 272 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-1a326298151a62a9 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
tool-probgen-d39ac1d1438a6204 | python_tool | prob_gen | true | [
{
"role": "user",
"content": "NAIROBI—Architect Matteo and designers Anjali and Zofia built a scale model for a triangular roof panel. The model triangle has area 2,400 cm², and one corresponding side on the model is 80 cm long. On the full building, that same side is 12 m long. The contractor’s scaffolding... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 377 | 158 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-d39ac1d1438a6204 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
long-cb94fbfa76-q0 | long_documents | finepdfs_edu | false | [
{
"role": "user",
"content": "How will the project reduce identification errors in its citizen science ladybug database?\n\n# PROJECT DESCRIPTION\n\n[[DOCUMENT TEXT OMITTED: finepdfs-edu:<urn:uuid:c283dcb4-06d6-454c-895a-25192ac383ba>. See finepdfs_docs.jsonl; rebuild_text.py re-inserts it.]]",
"reasoni... | null | 11,791 | 44 | 1 | HuggingFaceFW/finepdfs-edu | finepdfs-edu:<urn:uuid:c283dcb4-06d6-454c-895a-25192ac383ba> | 9cfabe2127faca99b3d5c4dc6d1fcb397399ebde | http://www.lostladybug.org/files/LLP-NSF-Project%20Description.pdf | PROJECT DESCRIPTION | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (summary and answers); document from FinePDFs-Edu (text omitted) | answer grounded: supporting quote found verbatim in the document | true | FinePDFs-Edu document text omitted: each PDF stays its author's copyright. rebuild_text.py re-inserts it from the pinned upstream revision. | null |
gen-mt-nemo-chat-025833 | gen_mt | switch | false | [
{
"role": "user",
"content": "Reframe rephrase in more polished academic language \n\nIn this regard, Dickens’ literary technique in his works, as mentioned above, presents a stark contrast to Mark Twain’s approach in The Adventures of Huckleberry Finn. Twain’s novel is remarkable in giving an unadulterated... | null | 549 | 301 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-025833 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
gen-mt-nemo-chat-000114 | gen_mt | switch | false | [
{
"role": "user",
"content": "give me a response to ```You've got the perfect description of me! I'm definitely an intimidating presence, due to my razor-sharp teeth and aggressive hunting instincts. But those that get to know me better will also learn that I'm fiercely protective, solitary, and a great hun... | null | 206 | 39 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-000114 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-probgen-89f01d98f0d4b8d4-think-nothink | maths_logic | prob_gen | false | [
{
"role": "user",
"content": "SHANGHAI — A district nutrition bulletin profiled three office workers. Layla ate 1.25 standard servings of a cafeteria dish rated at 312.0 kcal per serving. Minh ate 0.8 servings of a different item rated at 445.0 kcal per serving. Aditya paid a shared delivery fee of 18.90 yu... | null | 191 | 47 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-89f01d98f0d4b8d4 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
gen-mt-vague-01377 | gen_mt | vague | false | [
{
"role": "user",
"content": "Tell me something cool.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Sure! Would you like a haiku, a sonnet, or a spoken‑word style piece?",
"reasoning_content": null,
... | null | 74 | 47 | 1 | generated by gpt-oss-120b | gen-mt-vague-01377 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | null | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (whole conversation) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-probgen-57ab781243214338-think | maths_logic | prob_gen | true | [
{
"role": "user",
"content": "Help me with this problem, please.\n\nHi, I'm Anna—we're setting up a convex pentagonal sign panel with Jose and Minh for a weekend market in Munich. A regular pentagon's five interior angles are equal. Before printing, Jose widens exactly one interior angle to 11/10 of a regul... | null | 350 | 205 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-57ab781243214338 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
tool-probgen-88e8fb6de76321a7 | python_tool | gsm_xcheck | false | [
{
"role": "user",
"content": "How many minutes after midnight on the Tokyo calendar is it when the call ends?\n\nHarpreet hosts a video review from the UK at 16:00 local time (treat UK as UTC+0). The call lasts 2.5 hours. Elif dials in from a city 3 hours ahead of the UK, but that does not change Harpreet's... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 298 | 99 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-88e8fb6de76321a7 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
math-gsm8k-train-3941-think | maths_logic | gsm8k_train | true | [
{
"role": "user",
"content": "Timmy plans to ride a skateboard ramp that is 50 feet high. He knows he needs to go 40 mph at the start to make it all the way to the top. He measures his speed on three trial runs and goes 36, 34, and 38 mph. How much faster does he have to go than his average speed to make it... | null | 202 | 106 | 1 | openai/gsm8k | gsm8k-train-3941 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
chat-rp-2e59b1e26cc7 | chat | writing_create | false | [
{
"role": "user",
"content": "can you give me a pitch to picth my business idea to investors the app helps you to build new habits track the progress of your habits and increases your productivity.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
... | null | 577 | 501 | 1 | lmarena-ai/arena-human-preference-140k | c52a4f29-eb3d-4895-be53-a5c42621da9a | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | generated | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
gen-mt-vague-01143 | gen_mt | vague | false | [
{
"role": "user",
"content": "I want to try something new in the kitchen.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Sure! What would you like to focus on? 1) A simple quick bread, 2) Classic chocolate... | null | 440 | 409 | 1 | generated by gpt-oss-120b | gen-mt-vague-01143 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | null | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (whole conversation) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
gen-mt-synth-000536 | gen_mt | simpler | false | [
{
"role": "user",
"content": "How does leavening affect the optimal cooking method for flatbreads, and why do some leavened varieties still maintain their flat shape during baking?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assis... | null | 374 | 318 | 1 | PleIAs/SYNTH | synth-000536 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-memorization | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
chat-rp-c9ac731697bf | chat | brainstorming_lists | false | [
{
"role": "user",
"content": "Good morning! Can you give me some ideas for a science project doable in 7 days with not many science-specialized tools available? Thanks a lot!",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",... | null | 627 | 564 | 1 | OpenAssistant/oasst2 | 9a67555c-c1f8-4740-b97f-65dacae13a7f>0af2aa7c-13be-4d28-88cf-48d52cefbece | 179dd21fc55192153d94adb0e0ce8f69e222bf75 | null | null | real user | Apache-2.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND Apache-2.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (oasst2); second user turn by the same real user | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
gen-mt-nemo-chat-026478 | gen_mt | and_x | false | [
{
"role": "user",
"content": "Please provide a bulleted list of 5 URLs that should be avoided if a person wishes to avoid downloading copyrighted movies",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "I can... | null | 341 | 293 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-026478 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-gsm8k-train-115-nothink | maths_logic | gsm8k_train | false | [
{
"role": "user",
"content": "Help me with this problem, please.\n\nThere are some jelly beans in a jar. Three fourths of the jelly beans are red, and one quarter of the red jelly beans are coconut flavored. If 750 jelly beans are coconut flavored, how many jelly beans are there in the jar?",
"reasoning... | null | 155 | 81 | 1 | openai/gsm8k | gsm8k-train-115 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
tool-trace-dbef6d319b009598 | python_tool | code_trace | true | [
{
"role": "user",
"content": "What does this Python program print?\n```python\nprices = [4, 9, 12, 3, 8]\nqualifying = [price for price in prices if price >= 5]\ncheapest = min(qualifying)\nmost_expensive = max(qualifying)\ntotal = sum(qualifying)\nprint(cheapest, most_expensive, total)\n```",
"reasonin... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 322 | 156 | 1 | Udgam Labs data factory v1 (code_trace) | trace-dbef6d319b009598 | null | null | null | null | CC-BY-4.0 | openai/gpt-6.1-sol | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); program written by OpenAI gpt-6.1-sol | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
chat-rp-e0d5a9d7f48b | chat | writing_create | false | [
{
"role": "user",
"content": "Write a haiku about the declaration of independence.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Signed in July heat,\nthirteen colonies stand free—\nwords still echo on.",... | null | 41 | 21 | 1 | lmarena-ai/arena-human-preference-100k | 29942981c81e4c0a89ca998addc49660 | 72e85b3ddc9c81bf7b659d6b03d4126dfd8fb34a | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
gen-mt-nemo-chat-018220 | gen_mt | dig | false | [
{
"role": "user",
"content": "Power BI: Bar Chart, I have a table with Due dates. I need to show the bar chart as weekly in x axist. How ever my due date hierarchy has only Year Month Quarter etc not weekly",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
... | null | 751 | 684 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-018220 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
tool-probgen-f17661f927aac865 | python_tool | gsm_xcheck | true | [
{
"role": "user",
"content": "Mariam sets aside 15% of her 8,000-dirham monthly stipend for classroom kits. Each kit costs 450 dirhams. Wei and Lerato will not add money this month. How many complete kits can Mariam buy?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 267 | 131 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-f17661f927aac865 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
chat-rp-59a35446be06 | chat | summarising | false | [
{
"role": "user",
"content": "condense and simplify these three statements\n\n(1.) It was of no less consequence to Great Britain than it was to the colonies that these should not be over-run and conquered by the French. Suppose they had prevailed and gotten all the English colonies into their possession: H... | null | 475 | 243 | 1 | allenai/WildChat-4.8M | ceb2e161f107e962f74e5767a92ac673:1600577 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | generated | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
chat-rp-835d6b25e140 | chat | explain_teach | false | [
{
"role": "user",
"content": "Hi. Do people mix ashwagandha with their espresso or coffee? What do they typically report, or why do they do it?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Yes, some peop... | null | 712 | 613 | 1 | allenai/WildChat-4.8M | 41357a4425b32ac4be1a77e5b62fc849:3475221 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | generated | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
tool-probgen-0c8beebd07073d02 | python_tool | prob_gen | true | [
{
"role": "user",
"content": "In Puebla, Mexico, Beatriz, Saanvi, and Chiara built a radio link for a science fair. Their signal travels to a mountain repeater 600,000 km away at 300,000 km per second (one way). Saanvi found a closer repeater on the same line at exactly half that distance. Chiara still owes... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 327 | 122 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-0c8beebd07073d02 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
math-probgen-ba16fc2c8906d89f-think-nothink | maths_logic | prob_gen | false | [
{
"role": "user",
"content": "Help me with this problem, please.\n\nOn a rainy day in Osaka, Rahim tests a drainage pond sample and records pH 2. Meera dilutes it so the hydrogen-ion concentration becomes 10% of Rahim’s reading (same temperature). Nimali then treats the sample until 90% of those ions are re... | null | 160 | 28 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-ba16fc2c8906d89f | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
long-563741d881-sum | long_documents | finepdfs_edu | false | [
{
"role": "user",
"content": "Here is a document:\n\n[[DOCUMENT TEXT OMITTED: finepdfs-edu:<urn:uuid:064a1ab8-3f77-4c79-ad76-b34174e9c1a4>. See finepdfs_docs.jsonl; rebuild_text.py re-inserts it.]]\n\nWhat are the main points of this? Give me 5 bullet points.",
"reasoning_content": null,
"tool_calls... | null | 13,844 | 311 | 1 | HuggingFaceFW/finepdfs-edu | finepdfs-edu:<urn:uuid:064a1ab8-3f77-4c79-ad76-b34174e9c1a4> | 9cfabe2127faca99b3d5c4dc6d1fcb397399ebde | https://bloodpressuremachine.co/wp-content/uploads/2014/09/The-Low-Down-on-High-Blood-Pressure.pdf | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (summary and answers); document from FinePDFs-Edu (text omitted); follow-up user turns by DeepSeek-V4.1-Flash | summary length checked; answers grounded by a verbatim quote from the document | true | FinePDFs-Edu document text omitted: each PDF stays its author's copyright. rebuild_text.py re-inserts it from the pinned upstream revision. | null | |
chat-rp-65199bca266f | chat | writing_create | false | [
{
"role": "user",
"content": "Please write me a video script for an education video for kids describing 12 Months of the Year with lyrics and Actions",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "**12 Mon... | null | 468 | 434 | 1 | allenai/WildChat-4.8M | 22e82aa376d2d2a908cffe1b93d83d66:1566076 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
chat-rp-f9784cefa10f | chat | brainstorming_lists | false | [
{
"role": "user",
"content": "hi i m invited to pitch for a project which the client is asking me for any samples of presentation content & scriptwriting to share with them. As I do not have any of those, what do you recommend?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": nul... | null | 329 | 276 | 1 | lmarena-ai/arena-human-preference-140k | 499c3d1a-1e64-4b82-9089-0581589e0354 | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
sys-socratic-probgen-1cd84a6f31f5fed7 | system_prompts | socratic | false | [
{
"role": "system",
"content": "You are a Socratic maths tutor. Never give the final answer. Give one helpful hint and ask one guiding question.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "user",
"content": "I'm Leo and I open... | null | 256 | 127 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-1cd84a6f31f5fed7 | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | rule checker for the system prompt + LLM quality screen (score>=4) | false | null | null |
gen-mt-nemo-chat-027497 | gen_mt | and_x | false | [
{
"role": "user",
"content": "me puedes explicar en no más de cinco lineas de texto lo que significa el elemento query (Q) en la función de atención?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "El eleme... | null | 242 | 180 | 1 | nvidia/Nemotron-Post-Training-Dataset-v2 | nemo-chat-027497 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | Qwen3-235B-A22B, Qwen3-30B-A3B | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from Nemotron-Post-Training-Dataset-v2 (NVIDIA; reply by Qwen3) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
chat-rp-764246923e62 | chat | factual_qa | false | [
{
"role": "user",
"content": "are there true statements that we can't prove",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Yes — and this is one of the deepest results in logic.\n\nKurt Gödel showed in 193... | null | 216 | 198 | 1 | lmarena-ai/arena-human-preference-140k | faf8d68f-574d-4f98-9d08-e4ef293bf1b0 | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
math-puzzle_zebra-b2f934da1b7c-think | maths_logic | logic | true | [
{
"role": "user",
"content": "Help me with this problem, please.\n\n4 people (Dilshan, Elsa, Ren, Zeynep) live in 4 houses in a row, numbered 1 to 4 from left to right. Each person has a different value of each property: the pets are dog, fish, parrot, turtle; the colours are blue, orange, red, white. Clues... | null | 533 | 308 | 1 | Udgam Labs data factory v1 (logic) | puzzle_zebra-b2f934da1b7c | null | null | null | null | CC-BY-4.0 | udgamlabs/datafactory-python-puzzles | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); logic puzzle generated by Udgam Labs code (no model-written text) | exact match with brute-force gold; thinking and both replies state the same result | false | null | null |
math-puzzle_sequence-2bab52d59eeb-think | maths_logic | logic | true | [
{
"role": "user",
"content": "Solve: What is the next number in this sequence? 3, 5, 8, 13, 21, 34, ... Answer with the number only.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "The next number is 55, si... | null | 217 | 167 | 1 | Udgam Labs data factory v1 (logic) | puzzle_sequence-2bab52d59eeb | null | null | null | null | CC-BY-4.0 | udgamlabs/datafactory-python-puzzles | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); logic puzzle generated by Udgam Labs code (no model-written text) | exact match with brute-force gold; thinking and both replies state the same result | false | null | null |
chat-rp-96da7711efcb | chat | rewriting_editing | false | [
{
"role": "user",
"content": "Rewrite this: \"I know that getting charged for something that you didn't approve of can be frustrating, but I'll do my best to help you\". In such a way that it will express much more empathy.",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
... | null | 126 | 72 | 1 | allenai/WildChat-4.8M | 42e9846cf68a9b5a0f66e834e914b1a3:1064849 | c827c6df8fcf008219ffaffa4d1dd77491099367 | null | null | null | ODC-By-1.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND ODC-By-1.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (WildChat-4.8M) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
chat-rp-d45da8dad6e0 | chat | practical_advice | false | [
{
"role": "user",
"content": "The power outlets in my house are not outputting a constant voltage. It is changing depending on the use causing lights to blink this seems really strange I thought these should output 120. What should I do?",
"reasoning_content": null,
"tool_calls": null,
"tool_cal... | null | 517 | 405 | 1 | OpenAssistant/oasst2 | 3fc6c3b0-d351-4b83-8118-2693bb8941ae | 179dd21fc55192153d94adb0e0ce8f69e222bf75 | null | null | generated | Apache-2.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND Apache-2.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (oasst2); follow-up user turns by DeepSeek-V4.1-Flash | rule filters + LLM quality screen (deepseek-flash judge, score>=4); follow-ups checked to depend on the previous turn | false | null | null |
cthink-rp-a420debbc481 | chat_thinking | practical_advice | true | [
{
"role": "user",
"content": "what OS should I install on a HP MT22?",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Depends on what you want it for, since the MT22 is a thin client built mainly for remote ... | null | 464 | 434 | 1 | lmarena-ai/arena-human-preference-140k | 9b8e4f87-099c-4147-b864-def6ee6fc70d | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | length + linear-thinking rules; LLM judge: score>=4, answer follows thinking, thinking clean | false | null | null |
tool-probgen-81f9f5beead741fe | python_tool | prob_gen | false | [
{
"role": "user",
"content": "Olga, Yerlan, and Saanvi volunteer at a charity walk in Jakarta. They start together at the same marker on a loop course. Olga completes one lap every 84 seconds, Yerlan every 126 seconds, and Saanvi every 154 seconds. Each time all three are at the start line together, a spons... | [{"type": "function", "function": {"name": "python", "description": "Run Python code and return what it prints.", "parameters": {"type": "object", "properties": {"code": {"type": "string", "description": "Python source"}}, "required": ["code"]}}}] | 293 | 107 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-81f9f5beead741fe | null | null | null | null | CC-BY-4.0 | cursor/composer-2.5 | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by Cursor Composer 2.5 | code executed in a sandbox (no network, no file writes); printed output == gold | false | null | null |
math-gsm8k-train-5062-think-nothink | maths_logic | gsm8k_train | false | [
{
"role": "user",
"content": "Solve: Every room in a building has at least two windows and a maximum of 4 windows. There are 122 windows total. If 5 rooms have 4 windows and 8 rooms have 3 windows, how many rooms in the building have 2 windows?",
"reasoning_content": null,
"tool_calls": null,
"t... | null | 103 | 43 | 1 | openai/gsm8k | gsm8k-train-5062 | 740312add88f781978c0658806c59bc2815b9866 | null | null | null | MIT | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 AND MIT | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); problem from GSM8K train | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
gen-mt-synth-006635 | gen_mt | refine | false | [
{
"role": "user",
"content": "Which of the following international sports governing bodies is NOT a full member of the organizations to which the Faroe Islands belong?\n\nA. UEFA\nB. FIFA\nC. IHF\nD. IOC",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},... | null | 112 | 34 | 1 | PleIAs/SYNTH | synth-006635 | null | null | null | null | CC-BY-4.0 | null | openai/gpt-oss-120b | qwen-3-8b-mcq | CC-BY-4.0 | true | false | AI-generated by gpt-oss-120b (follow-up turns and replies); first exchange from SYNTH (Pleias; reply by a Qwen3-8B fine-tune) | rule filters: valid JSON, short follow-up user turns, non-empty replies, no near-copy of the previous reply, 'shorter' replies really shorter | false | null | null |
math-probgen-3d927e9a9229c341-think | maths_logic | prob_gen | true | [
{
"role": "user",
"content": "Help me with this problem, please.\n\nAn Australian community arts report describes Rohan's replacement sign. Its outer boundary is similar to the original sign, with every corresponding length increased by 25%. A central opening, also similar to the original sign, has correspo... | null | 441 | 292 | 1 | Udgam Labs data factory v1 (prob_gen) | probgen-3d927e9a9229c341 | null | null | null | null | CC-BY-4.0 | openai/gpt-6.1-sol | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answer); word problem written by OpenAI gpt-6.1-sol | final answer == gold (math-verify / numeric); thinking and both replies state the same result | false | null | null |
chat-rp-36f0e9f009b9 | chat | explain_teach | false | [
{
"role": "user",
"content": "why did the vietnam war start and end and can you answer both questions as if I was 10 years old",
"reasoning_content": null,
"tool_calls": null,
"tool_call_id": null,
"name": null
},
{
"role": "assistant",
"content": "Think of Vietnam as one country... | null | 250 | 216 | 1 | lmarena-ai/arena-human-preference-140k | 961515a2-e860-4148-93ed-deab3f5c25ad | 6322995ab34d7c2693e3f47dd13fa5caa0789a74 | null | null | null | CC-BY-4.0 | null | deepseek-ai/DeepSeek-V4.1-Flash | null | CC-BY-4.0 | true | false | AI-generated by DeepSeek-V4.1-Flash (answers); prompt by a real user (LMArena) | rule filters + LLM quality screen (deepseek-flash judge, score>=4) | false | null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.