parent_id stringlengths 16 32 | id stringlengths 16 32 | split stringclasses 1
value | messages_json stringlengths 630 244k ⌀ | tools_json stringclasses 204
values | task_type stringclasses 62
values | source stringclasses 32
values | domain stringclasses 16
values | subdomain stringclasses 62
values | teacher_model stringclasses 3
values | teacher_provider stringclasses 4
values | trace_kind stringclasses 3
values | disposition stringclasses 4
values | ground_truth_json stringlengths 1 26.1k ⌀ | source_record_hash stringlengths 71 71 ⌀ | prompt_cluster_id stringlengths 24 24 ⌀ | template_cluster_id stringlengths 24 24 ⌀ | dedup_cluster_id stringlengths 24 24 ⌀ | verifier_passed bool 1
class | family_oracle_passed bool 1
class | reference_agreement bool 1
class | format_score_raw float64 1 10 ⌀ | quality_flags listlengths 0 3 ⌀ | sampling_weight float64 0.4 4.21 | n_assistant_turns int32 1 12 ⌀ | n_tool_calls int32 0 13 ⌀ | schema_version stringclasses 3
values | source_repository stringclasses 20
values | source_revision stringclasses 21
values | source_license stringclasses 11
values | source_split stringclasses 4
values | source_item_id stringlengths 16 41 ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
010d310f7fea2ad3db266027c31bc52c | 010d310f7fea2ad3db266027c31bc52c | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | NuminaMath-CoT | NuminaMath-CoT | math | NuminaMath-CoT | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "① Sampling from a uniformly moving production line, where a quality inspector takes one product every 10 minutes for a certain index test, is systematic sampling, not stratified sampling. Therefore, proposition ① is false.\n② According to the meaning of the linear correlation coefficient $r$, when two random variables... | sha256:ac2f2b10b97ffd983e02b9f6c0ed2227392819175c7bd1e03dcdec9ca77773fd | c521510dacab3aba251b7b8c | 2853fb3a97f2182cd685f98f | 4963a66c217798b28b8ed3a4 | true | null | true | 9 | [
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | AI-MO/NuminaMath-CoT | 9d8d210c9f6a36c8f3cd84045668c9b7800ef517 | Apache-2.0 | train | curated:cf009551a793b27e |
010d3a3626fd82147e5ab7c2865b357c | 010d3a3626fd82147e5ab7c2865b357c | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search a fixed local document corpus and return matching passages.","name":"corpus_search","parameters":{"properties":{"query":{"description":"Search query","type":"string"},"top_k":{"default":5,"description":"Number of results to return","type":"integer"}},"required":["query"],"type":"obje... | multi_research | glm5.2-agent-tool-synthetic | agent_tool | multi_research | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"multi_research","expected_final":"Cite the retrieved passages for latency=36 and batch size=240, then report the calculator result for 240 / 36 as 6.6667.","expected_tools":["corpus_search","corpus_search","corpus_search","calculator"],"seed":13000826} | sha256:7520e8c80b3aa42f9cf384faf7a7c7cc22e0d540de2cb92117358fef927fd2c7 | c2e9b4e8a32718e67d43bac6 | 6dc23b02eaa57d945bd81a70 | 85ac8e6fa4003b2715e8f234 | true | true | null | null | [] | 1.611823 | 4 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-multi_research-13000826 |
010daedc9cb95d043fe6ae747af159d5 | 010daedc9cb95d043fe6ae747af159d5 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search a fixed local document corpus and return matching passages.","name":"corpus_search","parameters":{"properties":{"query":{"description":"Search query","type":"string"},"top_k":{"default":5,"description":"Number of results to return","type":"integer"}},"required":["query"],"type":"obje... | multi_research | glm5.2-agent-tool-synthetic | agent_tool | multi_research | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"multi_research","expected_final":"Cite the retrieved passages for error rate=48 and batch size=480, then report the calculator result for 480 / 48 as 10.0000.","expected_tools":["corpus_search","corpus_search","calculator"],"seed":13001256} | sha256:064a3f3741089db8aa8b258fe285bdca7e958c08f6100c66fe649fc22c8b71b7 | ce1e5d5b30db05731d6e18c0 | b9c0d37585bd0df447d34632 | 9487ae937da48654db044aa3 | true | true | null | null | [] | 1.611823 | 3 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-multi_research-13001256 |
010e8b2dd733432ebda32f43b8cc626e | 010e8b2dd733432ebda32f43b8cc626e | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "If Minyoung arrived 33rd and Eunji arrived 11 places after Minyoung, you would add 11 to Minyoung's rank to find Eunji's rank.\n\n33 (Minyoung's rank) + 11 (the number of places after Minyoung that Eunji arrived) = 44\n\nSo, Eunji's rank is 44th." | sha256:9d921d2062f0c0c97de6aaf2dd6ea0bd560022126307747088dbdef1e468a9f1 | 271996fd17682d3474b0b51d | d995eac7723b1d40171edaea | 18fbeadd4e7b1cf0bf9ada61 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:352ae9a70a7220ed |
0111bf1001765eb4c0b514edd14cb175 | 0111bf1001765eb4c0b514edd14cb175 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here's a Python function that uses regular expressions to match the given criteria:\n\n```python\nimport re\n\ndef match_strings(string1, string2):\n pattern1 = r'^[a-zA-Z]+$'\n pattern2 = r'^(?=.*\\d)[a-zA-Z0-9!@#$%^&*()-=_+]+$'\n \n if re.match(pattern1, string1) and re.match(pattern2, string2):\n ... | sha256:7ec0285e83ec17ab76d798bd851e70ee7267737940c8b993a556efd2e6c5c970 | 7f5b69448ac7914cc5869c44 | 07bc25d4f8b3427dd88dc2d9 | 0e2a80d6a84b47d797de253b | null | null | null | 8 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:8d5654311c30ba3e |
011516c7be8e49ce2444a6a427925aaa | 011516c7be8e49ce2444a6a427925aaa | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Jacob ate a total of 400 + 900 + 1100 = 2400 calories.\nHe planned to eat less than 1800 calories.\nSo, he ate 2400 - 1800 = 600 more calories than he planned.\n#### 600\nThe answer is: 600" | sha256:a9f5732453a7088c6332afa22adb674d34a2e40e23a1c4fc6aa4757b81b6fdc5 | 3f44d19f578612a40f6b8bda | d4a57c0894a26ed817430d31 | 76e3969e3f876a9e9c8389cb | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:ea1668dffe9d0b31 |
01152ee887c26a020873bee010021c7c | 01152ee887c26a020873bee010021c7c | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "e hundreds digit, 3 choices for the tens digit, and 2 choices for the units digit, for a total of $4 \\times 3 \\times 2 = \\boxed{24}$ distinct three-digit numbers.\nThe answer is: 24" | sha256:f2e8f0771eb9b363df117b3704115bee175cbf6810d33328a63110d424a11b9e | 5825ece502bdea580add5589 | b8c5a51eefc5bf994af0c419 | 87206a6b3cfbbe214f033c69 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:03e2cec6208a25cb |
0117246963e71db7eee32770db10585b | 0117246963e71db7eee32770db10585b | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | CommonsenseQA | CommonsenseQA | reasoning | CommonsenseQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "E" | sha256:94111b1b22c43d8fc749ca10a81eef1aba185b5479e43425f3215c965b768169 | 20ea52c8a7eee78943958363 | 591634ff6bc534d120372af5 | e09d8d44c8f4fb3c710659f3 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | tau/commonsense_qa | 94630fe30dad47192a8546eb75f094926d47e155 | unknown-see-provenance | train | curated:ecb552a0b20da0f6 |
0118e296225430999e13d71270adb5a6 | 0118e296225430999e13d71270adb5a6 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | NuminaMath-CoT | NuminaMath-CoT | math | NuminaMath-CoT | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Let's call the two numbers x and y. We are given two equations:\n\n1) x + y = 22\n2) x^2 + y^2 = 404\n\nWe want to find the product of the two numbers, which is xy.\n\nFrom equation 1, we can express y in terms of x:\n\ny = 22 - x\n\nNow, let's substitute y in equation 2 with the expression we found from equation 1:\n... | sha256:cd7fe02a6a04e7263365de3c6696c981c8fdb63e6db3f1ff12ef12afc0c9cc3a | ff3af7c6929f6bc78c503a9a | 336bbb5745df9a8c5b567df3 | 3577daa4304bdbb25364ca52 | true | null | true | 9 | [
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | AI-MO/NuminaMath-CoT | 9d8d210c9f6a36c8f3cd84045668c9b7800ef517 | Apache-2.0 | train | curated:1e5b434a111e3538 |
011987abcc8f793cb4cc232347ee96dd | 011987abcc8f793cb4cc232347ee96dd | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | Dolly | Dolly | instruction | Dolly | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Pickleball is a slower-paced game. The court is smaller than a tennis court so there is less ground to cover. Pickleball paddles are lighter than tennis rackets; even the ball itself is lighter. Pickleball is typically easier on your joints compared to tennis. Some believe pickleball is an easier sport to learn." | sha256:c1e4f449d2cf93f0bccf568484f97863b70772e0537331111ecda8643b368693 | ae613b68b0fc65f8c15750df | 70b1a6a52e24da6d8407dc50 | 4c3f4e3617ea1144127eabc5 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | databricks/databricks-dolly-15k | bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a | CC-BY-SA-3.0 | train | curated:78653a2eabc7ab45 |
011a5694059ea95fa75a6f4929910798 | 011a5694059ea95fa75a6f4929910798 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "If Alaya's restaurant has 200 fruit salads, then Angel's restaurant would have 200 x 2 = <<200*2=400>>400 fruit salads.\nThe total number of fruit salads in both restaurants is 200 + 400 = <<200+400=600>>600.\n#### 600\nThe answer is: 600" | sha256:bbd9bd32d03b3515bbf2b5084d2cc5d22cf39f2eb877f73aeb758de1f88acf71 | ac5a67ff36ac05f131a43ce8 | 32c422f1d05fba7bcf66a63c | 5094073c339de97904a2cc27 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:dcd06b4a220675b2 |
011b5502ad25c8448337818d95d2383a | 011b5502ad25c8448337818d95d2383a | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Execute a SQL query against a SQLite database and return results.","name":"sqlite_query","parameters":{"properties":{"database":{"description":"Path to SQLite database file","type":"string"},"query":{"description":"SQL query to execute","type":"string"}},"required":["database","query"],"typ... | sqlite | glm5.2-agent-tool-synthetic | agent_tool | sqlite | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"sqlite","expected_final":"After the repair, the final query must verify reading=10 and quality=review for row 30 in measurements.","expected_tools":["sqlite_query","sqlite_query","sqlite_query"],"seed":14000599} | sha256:1793d45e47e9870b57892a38feb918fbc18413394ffba857bed110c2644130df | f71423fd8ba27a11c916e004 | e6539d65d99acd955f24b933 | add3f74a60950808435402b7 | true | true | null | null | [] | 1.611823 | 4 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-sqlite-14000599 |
0121e58d10298574f5bcc8ecb5a07d90 | 0121e58d10298574f5bcc8ecb5a07d90 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | CodeAlpaca | CodeAlpaca | code | CodeAlpaca | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "The Python built-in function to swap two values is the \"tuple unpacking\" operator (x, y = y, x). This allows you to swap two variables without having to use a temporary variable." | sha256:49e039c359d407d41e7a91052fdbb37b14b5fea8e4d80da70df2dbc39bd7de7c | 56659309ca63e734767d04e7 | b4f05b7e4e6445813f844a6a | ff003cc8455dcf8594a0f83c | null | null | null | 7 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | sahil2801/CodeAlpaca-20k | 152bb5e9a29651266b018106053980070a0521a1 | unknown-see-provenance | train | curated:82ee46841466cd72 |
0122f3ec13af892e0c575e92da046b2d | 0122f3ec13af892e0c575e92da046b2d | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here is an updated version of the code that incorporates the mentioned requirements:\n\n```python\nfruits = ['apple', 'banana', 'cherry', 'dragonfruit']\nvalid_fruits = ['apple', 'banana', 'cherry', 'dragonfruit']\n\ncounter = 0\n\ntry:\n for item in fruits:\n if isinstance(item, str):\n if item i... | sha256:ab604c529fdf667b115e823f94c82f3b65c1881b88484497ea7be9f828e2a75f | c2448228ced2aae000c493a2 | e9af5b856b82f22dae03cd86 | fd9a31632e5de2e2cfe61d55 | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:2b3dcc221616adee |
0124a4b21d32ca5b255d17316ee40737 | 0124a4b21d32ca5b255d17316ee40737 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search for files matching a pattern or containing a text pattern.","name":"search_files","parameters":{"properties":{"path":{"default":".","description":"Directory to search in","type":"string"},"pattern":{"description":"Search pattern (glob or text)","type":"string"}},"required":["pattern"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 3' with 'RETRY_LIMIT = 5' in /tmp/agent_tool_workspace_10000337/src/parser.py, and read the file back to confirm the exact edit.","expected_tools":["search_files","read_file","patch_file","read_file"],"seed":100003... | sha256:0660d4e3edae80f953a61069aa1e4fcd51a778664039295c015cb1805da06435 | 6105b15358ced99608df59d0 | 5b8a450ada3c9ddc96f151bf | 62085a10aa8a41d9c04c5a76 | true | true | null | null | [] | 1.611823 | 4 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10000337 |
0128710a7067dc9ff9053ec8542b1102 | 0128710a7067dc9ff9053ec8542b1102 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here is a Python program that takes a list of integers as input and returns the sum, count, and average of all even numbers in the list:\n\n```python\ndef calculate_even_sum(lst):\n count = 0\n even_sum = 0\n\n i = 0\n while i < len(lst):\n if lst[i] % 2 == 0:\n count += 1\n ev... | sha256:641bbaa389a288bbb57073902d3937cae253481e8845620ae3678644f8ed219f | 35884a8c7bde87e4f2da6541 | 75b4a36d3b0fef4599c3bb87 | 7cc70010507d2a102a799af9 | null | null | null | 8 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:38c4b7c7f772487e |
01299a8fd73eeed639ecbec8d907ae4f | 01299a8fd73eeed639ecbec8d907ae4f | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | Dolly | Dolly | instruction | Dolly | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Traditionally, banks accept deposits and loan that money to borrowers providing a time arbitrage on the velocity of money which regulates in profit. In more recent years, banks have diversified their profit centers by investing those deposits in fixed income securities. There has even been a proliferation of into mult... | sha256:d57069e4099c7ef45e2ec53b2a78e470f999a39c99bf6b6c5e83bb1865d56d68 | 19c37750f614eb705d7d732e | 8ca401418a56de3622734937 | 5b0c20eedfe946edede9f164 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | databricks/databricks-dolly-15k | bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a | CC-BY-SA-3.0 | train | curated:bca26321bdc7c376 |
012af6b7cdf736bfb0d702ccbe2f1e18 | 012af6b7cdf736bfb0d702ccbe2f1e18 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "A dozen is equal to 12, so 2 dozen eggs is equal to 2 * 12 = 24 eggs.\nIf Megan dropped a tray of eggs, she lost 3 eggs.\nShe also had twice as many cracked eggs, so she had 3 * 2 = 6 cracked eggs.\nThe total number of broken and cracked eggs is 3 + 6 = 9 eggs.\nTherefore, the number of eggs that are still in perfect ... | sha256:94493c9428f42ee0a2353d02a18d63351186cba11998de58a593f0439da6b84f | a80ea650712a6ad5f3ffa219 | 95259f51b0b548bf0fdbe641 | c32b55b03ba8a30248d8fdba | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:5133b9a35a7f2e74 |
012d66b7b0193bf203960f4123fe8a68 | 012d66b7b0193bf203960f4123fe8a68 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | NuminaMath-CoT | NuminaMath-CoT | math | NuminaMath-CoT | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Let $Q$ be the point on the unit circle that is $240^\\circ$ counterclockwise from $(1,0)$. To find $\\cos 240^\\circ$, we analyze the position of $Q$ on the unit circle. \n\n- **Step 1**: Determine the quadrant where $240^\\circ$ lands. Since $240^\\circ = 180^\\circ + 60^\\circ$, $Q$ lies in the third quadrant where... | sha256:3e62da6579db5893a3c29b3de91692b287a2b0521005f4e81fd69d025dd21245 | 778a51ef07113bc5a6289965 | f60e41e3b0a4f64fff65f646 | bab5ff68240c0cf0eb628899 | true | null | true | 9 | [
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | AI-MO/NuminaMath-CoT | 9d8d210c9f6a36c8f3cd84045668c9b7800ef517 | Apache-2.0 | train | curated:2de49d8f3e5dded6 |
012d9cc057789aca23971366aa0801ba | 012d9cc057789aca23971366aa0801ba | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "public class Rectangle {\n private double width;\n private double length;\n\n public double calculateArea() {\n if (width < 0 || length < 0) {\n throw new IllegalArgumentException(\"Width and length must be positive numbers\");\n }\n \n if (width == 0 || length == 0) {\n return 0;\n }\n ... | sha256:16ff8ea2e03edf4f206a87af61b64aa6d3cbc200cd9722a34429fb0104b58610 | 489571a28c1ec3551c1f6a64 | 999557c4b87295141ac12c58 | e270af7c3fbc5f067de7bc90 | null | null | null | 7 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:36d60bad4c381ac2 |
012fe21c3bbb31fcf9813335cb235217 | 012fe21c3bbb31fcf9813335cb235217 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | QASC | QASC | reasoning | QASC | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "D" | sha256:5973ad3fde709bed0eb27f346515cd99a877f0a94ef5ec2f9e6d795010c2e41f | e9f861f2419e82d2d321712a | 2e70c12c4b4d4b16e1f15e20 | fb76c9c35645600d264bb811 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/qasc | a34ba204eb9a33b919c10cc08f4f1c8dae5ec070 | CC-BY-4.0 | train | curated:be28c4e0c78280e6 |
013032de30f06de50a9e6dd6bf76e3aa | 013032de30f06de50a9e6dd6bf76e3aa | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 4' with 'RETRY_LIMIT = 6' in /tmp/agent_tool_workspace_10002060/src/parser.py, and read the file back to confirm the exact edit.","expected_tools":["read_file","search_files","search_files","read_file","patch_file"... | sha256:9a38ebd15825970740bf2a383bcb558a0fb7b61469f2fa686aa741d386c3a2ea | 5afbdf88ac0540e8c30daefc | 0ba048f17fa9730b777fa128 | f3cd1aecca622f3e59e1bdb2 | true | true | null | null | [] | 1.611823 | 4 | 6 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10002060 |
01309f973fb610688d0a7a73ecbc7f1b | 01309f973fb610688d0a7a73ecbc7f1b | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To solve this problem, we can use dynamic programming to find the optimal sum. Here is the algorithm:\n\n1. Initialize an array dp of size target + 1, where dp[i] represents the minimum number of unique elements required to reach the sum i. Initialize all elements of dp to infinity, except dp[0] which is set to 0.\n\n... | sha256:b1e56f187a5e5be30fee7b20ff6d55b5977bd960889f6a7ef5e3690d6f0a9671 | be2411c544ca5cdebd33beff | 84d048a6482d1e0a2d3754ef | c8f10a9ecd6e7589ef28e6ba | null | null | null | 8 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:0cad5b98334ca3c9 |
0130aa95966df69e1d94b75b361f8b02 | 0130aa95966df69e1d94b75b361f8b02 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To solve this problem, we need to determine the value of x, which represents the number of times Quincy sold as many stuffed animals as Thor.\nLet's break down the information given:\nNumber of stuffed animals sold by Jake: Thor + 10\nNumber of stuffed animals sold by Quincy: x * (Thor + 10)\nNumber of stuffed animals... | sha256:0bd663383c64c0e67fd461edbf756c032eff291cea7fecf874f18826f05828e9 | 5fc5bdc634fb0a5963cedacd | 29f8113861c89b8df7b23bb5 | 4fec58800d62e90bfb59d44c | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:f03b236f5c8c8cf3 |
01322bdf4cd698c23ef71a43005b937f | 01322bdf4cd698c23ef71a43005b937f | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To find the area of the connected paper, we first need to determine the total length of the paper once they are connected in a single line.\n\nSince each square-shaped piece of paper has a side measuring 30 cm, and Sumin overlaps them by 7 cm to connect them, the effective length added by each subsequent piece of pape... | sha256:51a6faaade0557e93bb622d4cf1729fb9f150a97b466903a7de406e07dfba667 | 1a7ab1133b82292ba793dc2c | c77170e57593aae1232ec5c9 | 37eecb253fa8e098dc5c92ec | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:c3c222e0bd101f8a |
013337a816650789c9481d86dca29a4d | 013337a816650789c9481d86dca29a4d | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Since order doesn't matter and we can't repeat friends, we need to find the number of combinations of 3 friends that we can choose from a group of 8.\nThis is given by the binomial coefficient $\\binom{8}{3}$, which can be computed as $\\frac{8!}{3!(8-3)!} = \\frac{8!}{3!5!} = \\frac{8 \\cdot 7 \\cdot 6}{3 \\cdot 2 \\... | sha256:c40a23c54864444167c5149be497e0f76c13577ffcd55276f59d20911808f232 | f0ade68f6b99d30caebb6856 | 41c702dfb235fd1f4b7770dd | 7db0f4025bba2e6f8e70cc43 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:b169e86c9598ff2d |
01333bf4805e3a18cd944202bda09677 | 01333bf4805e3a18cd944202bda09677 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "During the 20 minutes that Polly is happy, she will tweet 18 times per minute, so 20 minutes x 18 tweets per minute = 360 tweets.\nDuring the 20 minutes that Polly is hungry, she will tweet 4 times per minute, so 20 minutes x 4 tweets per minute = 80 tweets.\nDuring the 20 minutes that Polly watches her reflection, sh... | sha256:2960fb798158f03a7949afd9c0021acdaf7e73770a4b1cc5897a2a8bf9869c26 | 0f2bd4143819ddaef18acaab | 5adedda424e2da5061e2c953 | 32c18f59ebd689dfc4707180 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:c5ad052d5cd5f2ca |
01333f0a1b3c5882c0a7cab2ea1f66f8 | 01333f0a1b3c5882c0a7cab2ea1f66f8 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | SciQ | SciQ | reasoning | SciQ | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "C" | sha256:c37aaa8182299020a5fdc10a08b7fbcc8810c63dd8a5e92d48f77e77e74da119 | 368c616b3edb4221743cd27d | 50964c5d7ca50a424fdfda6b | b38a5d2451a3984d2b7de95f | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/sciq | 2c94ad3e1aafab77146f384e23536f97a4849815 | CC-BY-NC-3.0 | train | curated:9e9161a93ec05af7 |
0139b818c1e65b6990dea07fed62d192 | 0139b818c1e65b6990dea07fed62d192 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To find the number that is divisible by both 2 and 5, we need to look for a number that is even (divisible by 2) and ends with a 0 or 5 (divisible by 5).\n\nAmong the given numbers, 16, 18, 20, and 25, the number 20 is even and ends with a 0, which means it is divisible by both 2 and 5. Additionally, 20 is greater tha... | sha256:e5d75ba941845fb5a11bac09070e78644b3bf295b76397e0f21269db3a96de88 | c8d92aa6f0857a77d0d79b3f | 41282ebc4c0cc9073380d67a | 9171fe623b4e005ec6a9c4d6 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:1975a56564cfb9f9 |
0139e0a1e82581420137c5cc4633140e | 0139e0a1e82581420137c5cc4633140e | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | NuminaMath-CoT | NuminaMath-CoT | math | NuminaMath-CoT | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Since $(2\\overrightarrow{BA} - 3\\overrightarrow{BC}) \\cdot \\overrightarrow{CB} = 0$, it follows that $(2\\overrightarrow{BA} - 3(\\overrightarrow{AC} - \\overrightarrow{AB})) \\cdot (\\overrightarrow{AB} - \\overrightarrow{AC}) = 0$,\n\nwhich simplifies to $(\\overrightarrow{AB} - 3\\overrightarrow{AC}) \\cdot (\\... | sha256:787595f13fe6ae398e775980e52e41910ea117397c49ea939276ae8607371b1e | c800e203e9bf8c6825128f51 | 675658000ebe2950b114b703 | 3c9d38a06b8b752d2485918e | true | null | true | 7.5 | [
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | AI-MO/NuminaMath-CoT | 9d8d210c9f6a36c8f3cd84045668c9b7800ef517 | Apache-2.0 | train | curated:40dca6a854306a5a |
013c61cc4270056587986a3a540e8524 | 013c61cc4270056587986a3a540e8524 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | CodeAlpaca | CodeAlpaca | code | CodeAlpaca | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "def sum_list(some_list):\n result = 0\n for num in some_list:\n result += num\n return result\n\nprint(sum_list(some_list)) # Prints 15.3" | sha256:5d6f8f55d2fa230264e7436d151cb6608c5fb128562bebc6a84bf32e2e6d710a | 8eca1c698baab3de972451d0 | 715864c905489a298d897ed4 | cfd017cb0d10e0625e1722f9 | null | null | null | 9 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | sahil2801/CodeAlpaca-20k | 152bb5e9a29651266b018106053980070a0521a1 | unknown-see-provenance | validation | curated:9460ac573fcd8bf7 |
013cd12e8295aeaea4c284f1a2337615 | 013cd12e8295aeaea4c284f1a2337615 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search for files matching a pattern or containing a text pattern.","name":"search_files","parameters":{"properties":{"path":{"default":".","description":"Directory to search in","type":"string"},"pattern":{"description":"Search pattern (glob or text)","type":"string"}},"required":["pattern"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 2' with 'RETRY_LIMIT = 5' in /tmp/agent_tool_workspace_10002296/src/settings.py, and read the file back to confirm the exact edit.","expected_tools":["search_files","read_file","patch_file","read_file"],"seed":1000... | sha256:0dfed63c74ca55e50229236c6ee758c09eeab8a7e2ea71e42f554ba339013fbd | 8cd2a7a44767ac195d876588 | 1f7cbb3637c13217c4e0c9c1 | 9ce38bdb7b9699e20d3bf1f4 | true | true | null | null | [] | 1.611823 | 4 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10002296 |
013da21615451ae523ce44747b00f0b9 | 013da21615451ae523ce44747b00f0b9 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | QASC | QASC | reasoning | QASC | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "E" | sha256:264e35650d0063784102690cb506d29f57a26167f7458947195457d34a671634 | ff74a60f9d27b9a38d9801b8 | 102d0e397b0ad5d4cb504c1e | 8f1c3fabed4780d7afb549f5 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/qasc | a34ba204eb9a33b919c10cc08f4f1c8dae5ec070 | CC-BY-4.0 | validation | curated:4277f56601905fee |
013ff21e6477b0d876b9e5ec3e290883 | 013ff21e6477b0d876b9e5ec3e290883 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Query or update a deterministic calendar/schedule. No live external actions.","name":"calendar_query","parameters":{"properties":{"action":{"description":"Action: query, add_event, update_event, delete_event, list_events","type":"string"},"date":{"description":"ISO date (YYYY-MM-DD)","type"... | planning | glm5.2-agent-tool-synthetic | agent_tool | planning | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"planning","expected_final":"Report the deterministic calendar results, then confirm the local maintenance window event only if the calendar tool reports success.","expected_tools":["calendar_query","calendar_query","calendar_query"],"seed":17001268} | sha256:01d86c2b8375ac7ea784bbeb3452b34bbc156ce4268dbf4672023a895d677470 | 231d3842bca17da7e92b7d7e | da2b9411107d775b6b1e6320 | 61bbfa856140b9c87127de4f | true | true | null | null | [] | 1.611823 | 3 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-planning-17001268 |
01415e1dd3bee38c1b5a619b0565a9c1 | 01415e1dd3bee38c1b5a619b0565a9c1 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "## Plan\n\n1. Create a function named `validateEmail` that takes in an argument named `email`.\n2. Create a variable named `regex` and set it equal to the regular expression `/^[^\\s@]+@[^\\s@]+\\.[^\\s@]+$/`.\n3. Return the result of `regex.test(email)`" | sha256:76bb0776aef542486eb78ae5989018cff3e68f6df8e175b13b670cc33e898639 | 5e123f2268b982167142c762 | e40124b5b35f5a88f17370d9 | 415406383251f31e1931d8ca | null | null | null | 8.5 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:d30c58b6d2f63ade |
0142e2be5941b74d58c5e4c87874871c | 0142e2be5941b74d58c5e4c87874871c | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | recovery | glm5.2-agent-tool-synthetic | agent_tool | recovery | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"recovery","expected_final":"Recover from the missing path, patch /tmp/agent_tool_workspace_12001008/config/pipeline.cfg from batch_size=8 to batch_size=16, read it back, and report the verified final line.","expected_tools":["read_file","search_files","read_file","patch_file","read_file"],"seed":12001008} | sha256:1c555f324fb9ce7184342a102c106164805856b75478988c2d881ffcf1b07e77 | 7a9136dccc955d3b5511a953 | 4222d5eaf7df44f7563f7e95 | e6c57972aa774d972a05076a | true | true | null | null | [] | 1.611823 | 5 | 5 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-recovery-12001008 |
0143c3c57ed77c8e6d85e27676dee4fc | 0143c3c57ed77c8e6d85e27676dee4fc | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MATH/algebra | MATH/algebra | math | MATH/algebra | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Since two of the lines that are the edges of the triangle are known, their intersection must be one of the vertices of the triangle. So we have $y=0$ (the $x$-axis) and $y=\\frac{2}{3}x+5$. Solving this equation, we find $0=\\frac{2}{3}x+5$, or $-5=\\frac{2}{3}x$, so $x=-\\frac{15}{2}$. Thus one of the vertices of the... | sha256:8b79c7dad946be381d4e73c56bd91ce3bba5b1a93521a0bc35c66563cbf5fb34 | 2c12b2eef6d903c308a729a2 | efdda6d753842a58c3a5aee8 | e1d8851b0db963b5026897f7 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | EleutherAI/hendrycks_math | 21a5633873b6a120296cce3e2df9d5550074f4a3 | unknown-see-provenance | train | curated:936d6165050ad59d |
01462c35d0d17b3c2eb9c633d95a4ce6 | 01462c35d0d17b3c2eb9c633d95a4ce6 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 2' with 'RETRY_LIMIT = 5' in /tmp/agent_tool_workspace_10002408/src/worker.py, and read the file back to confirm the exact edit.","expected_tools":["read_file","search_files","read_file","patch_file","read_file"],"... | sha256:630547762206759367123c76edfdd945ede4e809f6d5902fc1d46ffb79e4a324 | db5a36a26c698d084af9c34c | 0a864fa743b3501ff553045e | 66cee73893b2b84cf52d5560 | true | true | null | null | [] | 1.611823 | 4 | 5 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10002408 |
01483556d63e19e76bf6757bde5e01ea | 01483556d63e19e76bf6757bde5e01ea | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | ARC-Easy | ARC-Easy | reasoning | ARC-Easy | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "B" | sha256:7ecd1abe8f2d27ba4febaa474246c4367c5ecb6e04eadfd4f4e304cd1ca60599 | 881440bab084b2ea53d72f6d | 3624f7bdd675c9231bf1b05b | 0564a9d0fa7a57c3447c47bb | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/ai2_arc | 210d026faf9955653af8916fad021475a3f00453 | unknown-see-provenance | train | curated:a17e0f6da8c851ce |
0149a7ad74fd15cb8be5bbeb0196416b | 0149a7ad74fd15cb8be5bbeb0196416b | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here's an example code snippet that implements the above approach:\n\n```python\ntext = \"Hello! My name is John. What is your name?\"\n\n# Split the text into sentences using \".!?\" as delimiters\nsentences = text.split(\".!?\")\n\n# Iterate over each sentence and add the appropriate punctuation mark at the end\nfor... | sha256:62af851268d2637de2b843af28318a03aad6eded4acfedf4ba04b2dbcd89d027 | 791b834ded1d910fcc92af33 | 4e5214571f7ed67c0b96902f | 306d8dd925cbf452d3982567 | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:eb723652add8874c |
014ab93d1d739387419f3c20b8a6fb7b | 014ab93d1d739387419f3c20b8a6fb7b | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Query or update a deterministic calendar/schedule. No live external actions.","name":"calendar_query","parameters":{"properties":{"action":{"description":"Action: query, add_event, update_event, delete_event, list_events","type":"string"},"date":{"description":"ISO date (YYYY-MM-DD)","type"... | planning | glm5.2-agent-tool-synthetic | agent_tool | planning | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"planning","expected_final":"Report the deterministic calendar results, then confirm the local design review event only if the calendar tool reports success.","expected_tools":["calendar_query","calendar_query","calendar_query"],"seed":17000855} | sha256:73ec5146e4253edf6ceb2c67dd9d9119ba2de518889be1d263dde4bff51146d2 | ee4c32b08d40372e02fbafac | abada60b1356f9bfe2a8da20 | 85e1e7796defffe4057adf93 | true | true | null | null | [] | 1.611823 | 3 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-planning-17000855 |
015024a3f0ac30d1966c95ebc54d396b | 015024a3f0ac30d1966c95ebc54d396b | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "he trapezoid and let $x$ be the length of $PQ$. Since $RS$ is twice the length of $PQ$, we have $RS=2x$. The area of a trapezoid is given by $\\frac{1}{2}(b_1+b_2)h$, where $b_1$ and $b_2$ are the lengths of the bases. In this case, $b_1=x$ and $b_2=RS=2x$. Therefore, the area of the trapezoid $PQRS$ is $\\frac{1}{2}(... | sha256:c38d5409f754b4d059ca51ec056955bbf50ca2cf982cda62878a3513f76d9c7f | ae184394a87aae24ad499f82 | 9ab63b5c7a184df1542d3179 | 1abdaa0ba2277a4619ec3306 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:d8596a5915273915 |
0150b26e7ebe7ec47ecce8917b149d55 | 0150b26e7ebe7ec47ecce8917b149d55 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | CodeAlpaca | CodeAlpaca | code | CodeAlpaca | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "program convertToHex;\n \n // Declare a variable to store the number\n var number: integer;\n \n // Declare a variable to store the result\n var result: string;\n \n // Get the number from the user\n write('Enter the number to convert: ');\n readln(number);\n \n // Call the recursive function to convert to hex... | sha256:c51d254445c74c313dc893cd6eb065e975a0ce0c38ae4b3199b6ec0bfd186532 | e08f8e1b5315c071061c5764 | f92e91d2da6ef578382b779d | 570541ab94b31e91dd30650a | null | null | null | 7 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | sahil2801/CodeAlpaca-20k | 152bb5e9a29651266b018106053980070a0521a1 | unknown-see-provenance | train | curated:a3137edb1a8b26cf |
0150d3bc88c53cf84e4729212c3a51cc | 0150d3bc88c53cf84e4729212c3a51cc | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here's a function in Java that finds the largest prime number in an array efficiently:\n\n```java\npublic static int findLargestPrime(int[] arr) {\n int largestPrime = 2; // Initialize with the smallest prime number\n \n for (int num : arr) {\n if (isPrime(num) && num > largestPrime) {\n lar... | sha256:952be30b8fc57a82bb3123f7e938d1afc681b547ab1c517ca241d57ee2a7be20 | 90deff0e567502500898670b | 5a74df8cdf92d3381f6a464c | 02aa4eae0ee0d6df42f925f4 | null | null | null | 8.5 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:5e58a74b90b89e37 |
0151bd695a162310fe62ce553ff0d22e | 0151bd695a162310fe62ce553ff0d22e | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | code_debug | glm5.2-agent-tool-synthetic | agent_tool | code_debug | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"code_debug","expected_final":"The deterministic test must pass after correcting /tmp/agent_tool_workspace_11000885/src/windows.py so that group([2, 4, 6]) returns 37; report the test output and the corrected behavior.","expected_tools":["read_file","patch_file","run_python"],"seed":11000885} | sha256:80b279f65002a3e9f7ee8f060797ffb2a404064d6dda78a3fe8d6b9c06d07428 | 02ad7c15ebfa280f95a1149a | 06bb1119c88462a2cf54a6a1 | 2cf16da404613bed3eef2895 | true | true | null | null | [] | 1.174049 | 4 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-code_debug-11000885 |
01533c381623d0aeb7aa8a377fa3a45b | 01533c381623d0aeb7aa8a377fa3a45b | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | NuminaMath-CoT | NuminaMath-CoT | math | NuminaMath-CoT | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Let's call the smaller number \"x\". We are given that the difference between the larger number (2520) and the smaller number (x) is 2415. So we can write the following equation:\n\n2520 - x = 2415\n\nNow, let's solve for x:\n\nx = 2520 - 2415\nx = 105\n\nSo the smaller number is 105.\n\nNow, we are told that when the... | sha256:ea9a6aa9aa4af9320538a608f75f072fe78445893fb0324825507084df51a0d0 | ebb291bdeca961e32d454564 | dd83d9413d905b6fa16bf96d | 11eb59acab8020c7a58777f2 | true | null | true | 9 | [
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | AI-MO/NuminaMath-CoT | 9d8d210c9f6a36c8f3cd84045668c9b7800ef517 | Apache-2.0 | train | curated:ab7c23e41a17b68f |
0154d85c1b1ee9967c3f735bdf968060 | 0154d85c1b1ee9967c3f735bdf968060 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To solve this problem, we need to determine the value of x, which represents the unknown variable in the given equation.\nThe equation is:\n$w + 2 - 3w - 4 + 5w + x - 7w - 8 = -4$\nCombining like terms, we get:\n-4w - 10 + x = -4\nTo isolate x, we add 4w and 10 to both sides of the equation:\n-4w - 10 + 4w + 10 + x = ... | sha256:d1be3bd7defd65fb669c4cb40eabd0c9bbfb1313bec9ce6324ba23a5ec12b9b0 | 01b19b2e2a13deada249263e | 7a61e6ca400ae72456f73b46 | 3a7a1945986f78f7ed78de6f | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | validation | curated:b87d95411ccf67bb |
0156052b7e5378f0593b0fce9cb8b327 | 0156052b7e5378f0593b0fce9cb8b327 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | code_debug | glm5.2-agent-tool-synthetic | agent_tool | code_debug | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"code_debug","expected_final":"The deterministic test must pass after correcting /tmp/agent_tool_workspace_11002612/src/totals.py so that compute([2, 4, 6]) returns 39; report the test output and the corrected behavior.","expected_tools":["read_file","patch_file","run_python"],"seed":11002612} | sha256:b0c28281a346224ca91f01923e688946183eaa8803d53b5df58765af3b3750de | 3cc651dd225152d140454bee | 74568478df9af8ff77715189 | fbc3beded2f917bc8ff06a57 | true | true | null | null | [] | 1.174049 | 4 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-code_debug-11002612 |
01576de3cf60a87bc84654f29e2b08aa | 01576de3cf60a87bc84654f29e2b08aa | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | SciQ | SciQ | reasoning | SciQ | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "D" | sha256:10a47e03c1103899f02f7617ba3f4bdd0ad86cd49852c755ebbd0105cca6b42f | cbd4f6475ab3f2dbacd157fd | 27a952c48ce75ca0158e76b0 | 7dc1aa469356388a76eaffde | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/sciq | 2c94ad3e1aafab77146f384e23536f97a4849815 | CC-BY-NC-3.0 | train | curated:68faed5a34bd3e38 |
015897f5c83c54819a406bcf3fc44909 | 015897f5c83c54819a406bcf3fc44909 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "# Step 1: Prompt the user to enter the starting and ending numbers for the range\nstart = int(input(\"Enter the starting number for the range: \"))\nend = int(input(\"Enter the ending number for the range: \"))\n\n# Step 2: Validate the input to ensure that the starting number is less than the ending number\nif start ... | sha256:1f064c3dd8d33e36b5216f5bef940bc74199dfb2b8a3b4ae544b8afe72f2ff8c | dabfe3f3ddd15002df475542 | 4c2158834c9b98f037182515 | 638258ffb19ee21814e7eba8 | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:60cc3572ca0b7db2 |
0158b3fd0ce69b0be67eb0d46c1fe5eb | 0158b3fd0ce69b0be67eb0d46c1fe5eb | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here's a Python program that meets the requirements mentioned above:\n\n```python\nimport math\nimport random\nimport multiprocessing\n\ndef is_prime(n, cache={}):\n if n in cache:\n return cache[n]\n if n < 2:\n cache[n] = False\n return False\n for i in range(2, int(math.sqrt(n)) + 1):\... | sha256:7fd5bfe84ebddbfd063ca0724845c3656c83e013606adec0fae0825b4af06fbf | a5bf7de6377d9d75d883f6d0 | c4f3ff6fe0a8c336fddc8110 | 50aa9328897b4b13fbc1d60f | null | null | null | 7.5 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:f1354643b296e671 |
01598f18f5eb364e304cf9638e652289 | 01598f18f5eb364e304cf9638e652289 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | ARC-Easy | ARC-Easy | reasoning | ARC-Easy | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "D" | sha256:745f0b9d7049d896b4fcd9b567ba0fc8b646807776f6c72e901ceeecb054cdb5 | 3a244fecf89f47b22174dbf9 | d4296444f8be07bfddde5746 | 9c57af6bfe5525b8140cd048 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/ai2_arc | 210d026faf9955653af8916fad021475a3f00453 | unknown-see-provenance | train | curated:307234b43b5a6952 |
015c0f0426d431103383eb46d544db9c | 015c0f0426d431103383eb46d544db9c | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Unfortunately, it is not possible to construct a regex expression to identify prime numbers within a given string of digits, especially with the given conditions. Regex is a powerful tool for pattern matching in strings, but it is not suitable for prime number identification or mathematical operations.\n\nIdentifying ... | sha256:51dfab7c6ffe78a9544293a0d7269db128a4662b97dc371df2e48fc736ff9040 | 4d47cb032d7a1db72bdcf390 | 4fe22152f989f593e077e6a8 | a72c6b8894fc9b5901ea6a66 | null | null | null | 7.5 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:2696299ed4ae0dd3 |
015df44f5c7bc8e4546d4efbae275222 | 015df44f5c7bc8e4546d4efbae275222 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Let $x$ be the number of teachers who had neither high blood pressure nor heart trouble.\nUsing the principle of inclusion-exclusion, we have $x = 120 - (70 + 40 - 20)$.\nSimplifying, we find $x = 120 - 90 = 30$.\nTherefore, $\\frac{x}{120} \\times 100 = \\frac{30}{120} \\times 100 = \\boxed{25\\%}$ of the teachers su... | sha256:8dfee6fd00674ba3e7884e87d7ead7f5a9280a31fa3141586443caeb9a4e7341 | 632944de91cf106be357ebcf | e8a417b99d3214340eddde6a | 9e2bf4f1b1beb2ea1dc271d8 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:031691a7ad6a9984 |
015e210af2a9ae629ddbaf757a7b26e2 | 015e210af2a9ae629ddbaf757a7b26e2 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Louise already has 28 toys worth $10 each, so the total cost of those toys is 28 x $10 = $280.\nIf the total cost of all the toys is $580 and the cost of the toys she already has is $280, then the cost of the 20 teddy bears she added is $580 - $280 = $300.\nTo find the cost of each teddy bear, we divide the total cost... | sha256:4ebd7a0c07fc2d16ce5b98fee8a96293e52b69aceb340df8cbcc40cd9ec9832e | 3b4cdb4fb9e213dbf934299e | a23e976bd6308f434b2857a0 | 2154724378be13b7ac38d929 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:35d79d4184741565 |
015f5d174b9b43a45d8a4f2693c72056 | 015f5d174b9b43a45d8a4f2693c72056 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:9675b30024126aee33cff3cb9c72c1fcdfc70a0485ef88c2da975ea803ed2b6c | 353b0ce67a9d8963315c8304 | 8d01b514d7e6ea4e5bcb6f8a | 1a6231feb064e511d1590970 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:21c6fdaaccb71955 |
01622c48d920e64ce58db68408337ced | 01622c48d920e64ce58db68408337ced | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 2' with 'RETRY_LIMIT = 7' in /tmp/agent_tool_workspace_10000280/src/worker.py, and read the file back to confirm the exact edit.","expected_tools":["read_file","search_files","search_files","read_file","patch_file"... | sha256:9197280e7d6c936e83424f4cc380a04c82488a64d955981af2e68c30bae76180 | 648229189ead08bc5b0b87fd | c2f5016550561cd9b056e9f1 | 07177bf6be33d05bfe01b9b1 | true | true | null | null | [] | 1.611823 | 4 | 6 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10000280 |
01626ed53ac0306a593e754d53410504 | 01626ed53ac0306a593e754d53410504 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Query or update a deterministic calendar/schedule. No live external actions.","name":"calendar_query","parameters":{"properties":{"action":{"description":"Action: query, add_event, update_event, delete_event, list_events","type":"string"},"date":{"description":"ISO date (YYYY-MM-DD)","type"... | planning | glm5.2-agent-tool-synthetic | agent_tool | planning | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"planning","expected_final":"Report the deterministic calendar results, then confirm the local release rehearsal event only if the calendar tool reports success.","expected_tools":["calendar_query","calendar_query","calendar_query"],"seed":17000688} | sha256:3bfff5158cbd0557f6269db5fadc8fc0f12aa1b9a1a6c81c37bb829c3e26ca26 | 47f1db80966efc88975215e7 | 10df016993279edd2a8e812a | a8fd3c9a85659e9e42159f18 | true | true | null | null | [] | 1.611823 | 3 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-planning-17000688 |
01627e2815928e0cd555dfda0e20b9a9 | 01627e2815928e0cd555dfda0e20b9a9 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | CommonsenseQA | CommonsenseQA | reasoning | CommonsenseQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "C" | sha256:869624526cc0dc5d495a1fdc9efe2cae0afcffa67b77a3d262d65be2fa305822 | ddfdafe0ed59e4b19e2c793e | 17a1ea14235d4be9aad12bcb | a690fcc6821753b4e1254948 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | tau/commonsense_qa | 94630fe30dad47192a8546eb75f094926d47e155 | unknown-see-provenance | train | curated:f3adc70e7f8b64b1 |
016323ca1b316529624cb07663546733 | 016323ca1b316529624cb07663546733 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:53949e098ac279ea409b9006acdcb41aad207b936a7c88d5c2d221b21f554387 | 9ce39693807acc4814bdb37b | 12fa27beb529d0bd85830f84 | 5d3889740e72774d1656303c | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:8cda689f906698eb |
016393b8abf699a83f221eec76edaf7d | 016393b8abf699a83f221eec76edaf7d | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | CommonsenseQA | CommonsenseQA | reasoning | CommonsenseQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "D" | sha256:262e59f388444ca84804208c947f2ab9bdd754ee7a7287a59f9cb9f73f81cf51 | 258f7d04dd4d3a4c72785805 | 240b8c32c9c16117c2a1eaa4 | 8cc6a0ca4515aa81b1adec09 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | tau/commonsense_qa | 94630fe30dad47192a8546eb75f094926d47e155 | unknown-see-provenance | train | curated:e30905ebc974c3d1 |
0163fdd56b1b93d82b4949602619cee3 | 0163fdd56b1b93d82b4949602619cee3 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "First, let's find the original number by dividing 458.64 by 14:\n\n458.64 / 14 = 32.76\n\nNow, we have the original number, which is 32.76. Next, we need to divide this number by 17 and round it up to two decimal places:\n\n32.76 / 17 ≈ 1.9270588235294118\n\nRounded up to two decimal places, the value is:\n\n1.93" | sha256:c0895ec97aa3456b99838d358ed123645ec8b237db0bf68aaa407b1ce7452823 | 0737dfac8455cf1f4a039771 | 9ea2dd9bb91986f0f6833393 | 2696e84c7d1802fa25b7aae7 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:6703b78f9de182af |
016457c5eea79ab8a059a7225e608ec9 | 016457c5eea79ab8a059a7225e608ec9 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | NuminaMath-CoT | NuminaMath-CoT | math | NuminaMath-CoT | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "The sum of the squares of the first n natural numbers is given by the formula:\n\nSum = n(n + 1)(2n + 1) / 6\n\nFor the first 60 natural numbers, we have n = 60. Plugging this into the formula, we get:\n\nSum = 60(60 + 1)(2*60 + 1) / 6\nSum = 60(61)(121) / 6\nSum = 60 * 61 * 121 / 6\nSum = 10 * 61 * 121\nSum = 610 * 1... | sha256:683e69bbe8997901f906befb0c299f6c884e8109812e3c08a639144d53876500 | 075ee106888af96ca902ab66 | 5dad175aef3fa35d1082ff0d | f2f02f91215a5dc108dda2c3 | true | null | true | 9 | [
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | AI-MO/NuminaMath-CoT | 9d8d210c9f6a36c8f3cd84045668c9b7800ef517 | Apache-2.0 | train | curated:0afb94d58aabb0a8 |
01651adcdc684c57419962da7493795a | 01651adcdc684c57419962da7493795a | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search a fixed local document corpus and return matching passages.","name":"corpus_search","parameters":{"properties":{"query":{"description":"Search query","type":"string"},"top_k":{"default":5,"description":"Number of results to return","type":"integer"}},"required":["query"],"type":"obje... | multi_research | glm5.2-agent-tool-synthetic | agent_tool | multi_research | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"multi_research","expected_final":"Cite the retrieved passages for error rate=24 and latency=120, then report the calculator result for 120 / 24 as 5.0000.","expected_tools":["corpus_search","corpus_search","calculator"],"seed":13000251} | sha256:24a61c06a84aa8126d980707c108e3d88439ecf8df934e28c595be6b07c2e2be | f6394a8258d0c8b910a15f46 | d5600fa6a1055d40461d6814 | 1faff776641930595000f515 | true | true | null | null | [] | 1.611823 | 3 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-multi_research-13000251 |
0167e45dc8fcf04f82596c9db6a70b16 | 0167e45dc8fcf04f82596c9db6a70b16 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "The area of a parallelogram can be calculated using the formula:\n\nArea = Base × Height\n\nGiven that the area of the parallelogram is 44 cm² and the height is 11 cm, we can rearrange the formula to solve for the base:\n\nBase = Area / Height\n\nPlugging in the given values:\n\nBase = 44 cm² / 11 cm\n\nBase = 4 cm\n\... | sha256:c25ff9b600ff4d3828f98d680887603fd2bcbcd28729b700ba45269ec2ef888d | cf78414a51ce3b24ef182232 | 345353458c404b5d61bcc6cb | 510606c7c517800f33d537e8 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:60c32b6cf71c5ea6 |
016802838d921a89062214b704fea50e | 016802838d921a89062214b704fea50e | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | SciQ | SciQ | reasoning | SciQ | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "D" | sha256:88f600c6767a5865b5a230a6ead9c65c9e7f180c3076d45fcbc4aa042bbef445 | 8d62140bcb26967b26fbcd29 | 5028d86f0d611d5779c97d8d | d59bab9192b1dc3a724550c3 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/sciq | 2c94ad3e1aafab77146f384e23536f97a4849815 | CC-BY-NC-3.0 | train | curated:99fd6a96bac8611a |
0168484940b9d1c07470d8af21f3d6f8 | 0168484940b9d1c07470d8af21f3d6f8 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:68a573845b38b09b4778ad7a94205555df608e38b382bca468071e2257a05347 | fe0ebd3449aef1f52907a3e4 | 99e1f5bbb628d396cf4095ec | f1b7adc9623c6b6c256b8756 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:6c9f0aded2ce609e |
0168fb650302e21f9a80659d6b68e88f | 0168fb650302e21f9a80659d6b68e88f | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Expanding $(x+m)^2+8$, we get $x^2+2mx+m^2+8$.\nComparing this with the original quadratic $x^2+bx+44$, we see that $2m=b$ and $m^2+8=44$.\nSolving the second equation, we find $m^2=36$, so $m=6$ or $m=-6$.\nSince $b$ is a positive number, we have $b=2m=2(6)=\\boxed{12}$.\nThe answer is: 12" | sha256:cf5878f29cc68ad0cf1e4c40c71b68891bcb1d61f92c568ba91b55009510dbef | 6d8c13b56f9b21c2a57c0c3e | 5f756f2df6a6e28853332a93 | df5116fae0969d5cdf898f30 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:c673fbb90f671445 |
016b54378be49355e3b1bd1f730f2591 | 016b54378be49355e3b1bd1f730f2591 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Let's denote the number of litres of water transferred from the left bucket to the right bucket as \\( x \\).\n\nAfter transferring \\( x \\) litres from the left bucket to the right bucket, the left bucket will have \\( 2800 - x \\) litres left, and the right bucket will have \\( 1500 + x \\) litres.\n\nAccording to ... | sha256:50ff284301a53df722075845daf99b0252036cc6aa6b4b90b9087fe2bba6c2a2 | f79a3ea028abd5efd8080cc1 | 6ffa07d334172694371c9e9f | a57d9cc97426e34c8a09f27f | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:75abde03dc3b0450 |
016c700a5572522eb123097ae3061b91 | 016c700a5572522eb123097ae3061b91 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here's a program in Python that meets the requirements:\n\n```python\ndef validate_input(numbers):\n try:\n numbers = numbers.split()\n numbers = [int(number) for number in numbers]\n for number in numbers:\n if not (1 <= number <= 1000):\n raise ValueError\n except... | sha256:53ac49ae54235a323bdcfa045bbdc66ab7489ba22c1195128bd21dac97885e23 | 5018703389213e62e01d1964 | f50fdc697a0a549ac6f1a46d | ca5c9b30af974b6fb7d058ec | null | null | null | 8 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:143010e165cf7271 |
01713bc9ab7b340a136097600e75872b | 01713bc9ab7b340a136097600e75872b | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:400590781de918ce69e5749b78ff7111c1d1ecdb04b05b568610b32998bfe447 | 782e1380df2debfa5e5c313a | ce15e94979abca0218e8f0bc | bb518a8731e500d67c75878b | null | null | null | 5 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | test | curated:834a23dc82433d89 |
0172af191edc41ab20681b9804201329 | 0172af191edc41ab20681b9804201329 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_recovery | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 2' with 'RETRY_LIMIT = 6' in /tmp/agent_tool_workspace_10001487/src/settings.py, and read the file back to confirm the exact edit.","expected_tools":["read_file","search_files","read_file","patch_file","read_file"]... | sha256:9b9434939b9370e18901d843b5d2b6811cf2ac4956fc4eaa79acd19acfb6aa4e | 6218b581ec2b315c8bc03399 | e76c48fa9c7c7419c1c4fc5f | 30d2565f01f01630616b2dd0 | true | true | null | null | [] | 1.611823 | 4 | 5 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10001487 |
0172bc31ed4d9b2a21b40e28e72c29a1 | 0172bc31ed4d9b2a21b40e28e72c29a1 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To find the number of occurrences of a particular character in a given string, while ignoring any occurrences within parentheses, curly brackets, single or double quotation marks, you can use the following approach:\n\n1. Initialize a variable `count` to 0 to keep track of the number of occurrences.\n2. Initialize a v... | sha256:d89f48a5641c2f13b724f74e806c3b1d0b163da4340573e6f48a095472a1fc03 | a0419cbab6a6f8c2a3c35e09 | dbee9714ddbcacc88519cc45 | 3fbe3a20afa0c1ffbfce6140 | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:55ff9252a7e2b915 |
017377a48e22c09fe257d7e0d4247de3 | 017377a48e22c09fe257d7e0d4247de3 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "def reverse_string(string):\n if len(string) == 0:\n return \"\"\n else:\n return reverse_string(string[1:]) + string[0]\n\ndef count_vowels(string):\n vowels = ['a', 'e', 'i', 'o', 'u']\n count = 0\n for char in string:\n if char.lower() in vowels:\n count += 1\n retu... | sha256:1fc6eca4b6d74371742c1a88aa218fd47497a3d10f9bd6906deee9662aaeaebd | beec387c2f6eb68f4eae31bb | 6663c53cf720e70f1a7cd686 | 7330dbbe59dc8b19f1fffddc | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:bd5f5f7eff3a4da3 |
01794d48d695c4aba6685d303e27a9f4 | 01794d48d695c4aba6685d303e27a9f4 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "The distance between (2,1) and (3,4) can be found using the distance formula:\n$\\sqrt{(3-2)^2 + (4-1)^2} = \\sqrt{1+9} = \\sqrt{10}$.\nSince the distance between adjacent vertices of a square is equal to the length of a side,\nthe area of the square is $(\\sqrt{10})^2 = \\boxed{10}$.\nThe answer is: 10" | sha256:1515aee963707308db4f59ed6c2c7e660ad8b5ba1b54c99eb7fcae7bfea0c9cc | 6dd8ebc8a4b22bf297ab958f | 1474b497e5937e03ec90ec26 | 88ddc219078029b44bf43bc0 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:54e153d53e60d463 |
01853a6450539b3cd8908fd063acbfc6 | 01853a6450539b3cd8908fd063acbfc6 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:e25626c70b7c08f08e7fc975849a72299cf653974cd5302ab46ca9fc25e976b6 | 32f6e5aab2fd7b3293490119 | ce2c8e9df84730f56461ff73 | 9be0d8d5910261efff262a90 | null | null | null | 5 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:ac0e04cdbd0484ca |
01860a7eb751386034731864f4c11d90 | 01860a7eb751386034731864f4c11d90 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Here's a Python program that prints the 42nd prime number and calculates the Fibonacci sequence using recursion:\n\n```python\nimport math\n\ndef is_prime(num):\n if num < 2:\n return False\n for i in range(2, int(math.sqrt(num)) + 1):\n if num % i == 0:\n return False\n return True\n... | sha256:06893c91f1827604abb2c9e5c47daeb4ced58ea8ca99975b84de8728adbffe0f | 385bacb924812e32f0cb8acf | 1f1149106c185a144a3b8e0d | 4532b036421c216bf8e38b0c | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:a536c79445d7c851 |
0187339ee8553071ef9399663ca33561 | 0187339ee8553071ef9399663ca33561 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | Dolly | Dolly | instruction | Dolly | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "The name \"rosemary\" derives from Latin ros marinus." | sha256:49f7481bf393aeb552af62bb0382f7b5c0b5bcfc6b189275f1ef6fd817fddd51 | 35b682b3a0e9d8509cb86c1e | b84c91ae20d7732e47910572 | 4a2882147e91bc3efc3a38ec | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | databricks/databricks-dolly-15k | bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a | CC-BY-SA-3.0 | train | curated:675c6369ee280e4c |
0187c4c043355521feccb9ba144fe9b9 | 0187c4c043355521feccb9ba144fe9b9 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "def remove_vowels(string)\n string.gsub(/[aeiou]/i, '')\nend\n\ndef count_non_alphabetic_chars(string)\n string.scan(/[^a-zA-Z]/).count\nend\n\ninput_string = \"Hello$#@World!\"\nresult = remove_vowels(input_string)\ncount = count_non_alphabetic_chars(result)\n\nputs \"Modified string: #{result}\"\nputs \"Number of ... | sha256:18be17bd61c0314d80e7801f359c3486943ca1f05cd44e8052760c057a5fb42d | 98e06df5ba21060fba8afdb9 | fa845573dcf873436c65ff48 | 8c378768a2db6f18684c149d | null | null | null | 7 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:6ac4c57a7484fcfb |
01882f5900b670b502b2519a0ec1c227 | 01882f5900b670b502b2519a0ec1c227 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | CodeAlpaca | CodeAlpaca | code | CodeAlpaca | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "<!DOCTYPE html>\n<html>\n<body>\n\n<form action=\"action_login.php\" method=\"post\">\n Username: <input type=\"text\" name=\"username\">\n <input type=\"submit\" value=\"Submit\">\n</form> \n\n</body>\n</html>" | sha256:352244328f90f9800ea7ed7535cc24603526e370696f2b321fac31ff9a2f50e4 | 534765cb9cfeea9126392cdc | cc83cae38215a807aa8fde1e | a92ad1cd09ab7cfcac7d7a13 | null | null | null | 8.5 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | sahil2801/CodeAlpaca-20k | 152bb5e9a29651266b018106053980070a0521a1 | unknown-see-provenance | test | curated:8f8344003b0f5e34 |
01882f6afa5e87c1860c5bc673f1f8b8 | 01882f6afa5e87c1860c5bc673f1f8b8 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:9062ea0ed962c1285203300636eb5ce40f1a65b4ade484d8b8a24471bc8f7ead | 4c4ecba4ca9d6f6f8e427b6f | c08893cd314d8b5656e9c50e | 3bc594d44d87bf352dad6730 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:f1e02e95520367ab |
0188b6bdc8d2d10856c2a2ca1ca427db | 0188b6bdc8d2d10856c2a2ca1ca427db | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Since $f(x)$ is odd, we have $f(-x) = -f(x)$ for all $x$ in the domain of $f$.\nApplying the function again, we have $f(f(-x)) = f(-f(x)) = -f(f(x))$.\nApplying the function one more time, we have $f(f(f(-x))) = f(-f(f(x))) = -f(f(f(x)))$.\nTherefore, $f(f(f(x)))$ is $\\boxed{\\text{odd}}$.The answer is: \\text{odd}" | sha256:530aac6c33f41762c22078a83cf991838789497b58f1705c3b6a08c008059887 | 0067c3da5499d17eb5ff9657 | 9f61fe5015f4e0d1fc0ee9d2 | 302cdf666cf593c00942cb87 | true | null | true | 9 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:259b0e2cd6836dae |
0189e58f886e0b134f4891e09e9195ba | 0189e58f886e0b134f4891e09e9195ba | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "```python\nclass Person:\n def __init__(self, name, age, hobby, friends):\n self.name = name\n self.age = age\n self.hobby = hobby\n self.friends = friends\n \n def increase_age(self, amount):\n self.age += amount\n \n def add_friend(self, name):\n self.friends.... | sha256:5b08b000b11a40af85ec6c8474bc45827b2e9931bf45c3e3a4b10f07b3ca8004 | 660d98e2ae25d7b0e68dba4b | 19792339ab7e99f0172156cc | 6a4d9626aa44ff5f337c7424 | null | null | null | 8.5 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | train | curated:64fe379fb547abf1 |
018a8e7e07ba8840df7be57985fdf395 | 018a8e7e07ba8840df7be57985fdf395 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | Dolly | Dolly | instruction | Dolly | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Malibamat'so River is a river in northern Lesotho. Its origin is near to the South African border, where it drains the eastern slopes of the Maloti Range. It flows southwards past the village of Lejone, and eventually joins the Senqu River 5 kilometres (3.1 mi) northeast of Mohlanapeng.\n\nThe Malibamat'so forms the n... | sha256:925243d7e65748717b24e76ff7e200f827dab93560d7d9dc352e54745a60e08e | beae73ec60c9d7ff38e6dca6 | 23385e0a754ee7b344b49186 | 0c90d454ee0bf7acf118f781 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | databricks/databricks-dolly-15k | bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a | CC-BY-SA-3.0 | validation | curated:918a8885e0072eba |
018c79c53cbe6f27f6f7aaf342dc481c | 018c79c53cbe6f27f6f7aaf342dc481c | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | ARC-Challenge | ARC-Challenge | reasoning | ARC-Challenge | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "B" | sha256:c87f28743fe23c00e0d2214ecdd9f3e303e857684d4715744557269fbfa95ba0 | e1fa8870f9a63b69066b2546 | 6b0a7936a646ac8211080af6 | 1adc9e42bb5435b495f8025c | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/ai2_arc | 210d026faf9955653af8916fad021475a3f00453 | unknown-see-provenance | train | curated:c67c178b936fae52 |
018c7d5db495c948e6f0d7c255f55af6 | 018c7d5db495c948e6f0d7c255f55af6 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:6af5d4e1353c2d8c8bf485fa8d80f95a8f187acd522037c06f29e9a2252fa693 | 6d1713978b3be4e77182a51f | 78bc1e050eb25e0bcde63c89 | 48d3c3641517ed82923e5e72 | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:c9bb4f0b8b18d8a3 |
018d0f8a3590adb08f2e4ff2ac9400c8 | 018d0f8a3590adb08f2e4ff2ac9400c8 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Only one straight line can pass through two distinct points. This is because a straight line is defined by two points, and any two points determine a unique straight line." | sha256:cf170031f4333a4b51796808169e66c3624eaa035c1637eeee22a0a183966790 | 3ec418f5482cc01064671e7d | dea4c262e79f989cbe5c319b | f5399b2cdb3214332fde5872 | null | null | null | 8.5 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:eedc839791610f24 |
018d92defaa4b2b495000148c7248478 | 018d92defaa4b2b495000148c7248478 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | CommonsenseQA | CommonsenseQA | reasoning | CommonsenseQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "D" | sha256:3ce8019aef9cbb1c3c86e336af08e0555503d71352eb2c98ea1d7e6638157401 | 7448346315f2e9729015fefa | b9794b5b56664c445ea888f1 | bade7be63e1920bc6b18c4ac | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | tau/commonsense_qa | 94630fe30dad47192a8546eb75f094926d47e155 | unknown-see-provenance | train | curated:805cb96f23f13819 |
018e2fa708ac849a47223ea0344bb5ff | 018e2fa708ac849a47223ea0344bb5ff | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To find the sum of the natural numbers in the range of 6 and 21, we can use the formula for the sum of an arithmetic series:\n\nSum = n/2 * (first term + last term)\n\nwhere n is the number of terms.\n\nFirst, we need to find the number of terms in the range from 6 to 21. Since both 6 and 21 are inclusive, we can calc... | sha256:28898907ee3efa56937d11f6bfc79596d804e309d3414008cd646db176b06b1b | d6371c716fdbb0dbdc8bdb7f | aef7abb1e2af7d5ed05ef0d0 | 2b98a7d008175eff5b7eb325 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:3868fad724ab6c84 |
018e81c1db9f3eab36988edf3fdf6f65 | 018e81c1db9f3eab36988edf3fdf6f65 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | CodeAlpaca | CodeAlpaca | code | CodeAlpaca | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "<ul>\n <li>3</li>\n <li>12</li>\n <li>13</li>\n</ul>" | sha256:d855a3bd0a868f99da4cb83ef0106b69be7214af1df07d8e3a2c435cdaf6fe8e | aeafc1c4ebf5a8e187d8ef4d | 86c58fed49b3ce5bc8647c2b | 91fe0debb72eb001fefef600 | null | null | null | 8.5 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | sahil2801/CodeAlpaca-20k | 152bb5e9a29651266b018106053980070a0521a1 | unknown-see-provenance | train | curated:e36a3e09c808cb21 |
018ef77e73a5b02365a3a18701ad8185 | 018ef77e73a5b02365a3a18701ad8185 | train | [{"content":"You are an expert software engineer. Write complete, correct, runnable code in the language requested by the user. Follow the requested language, interface, and constraints. Explain the approach briefly when requested.","name":"","reasoning_content":"","role":"system","tool_call_id":"","tool_calls":[],"tra... | [] | Evol-Code | Evol-Code | code | Evol-Code | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "import re\nfrom nltk.corpus import stopwords\n\ndef tokenize_and_remove_punctuation(sentence):\n # Tokenize the sentence into words\n words = re.findall(r'\\b\\w+\\b', sentence)\n\n # Remove any punctuation marks from the words\n words = [word for word in words if word.isalpha()]\n\n # Remove any stop w... | sha256:2d10501f072669407714ec448a9095814773cbdb4fd0ba9fb9df2e1efb024e73 | cc4a4a8d90c4576cbe2f51e9 | 4f15005031d7d5b902aa2f3a | 482dddd7babf482136cd428b | null | null | null | 9 | [
"difficulty_untrusted",
"reference_status_unknown",
"reference_unknown"
] | 0.395189 | 1 | 0 | 1.0.0 | nickrosh/Evol-Instruct-Code-80k-v1 | 3ae930c20d5496e2c8386872d5628c45f6957db4 | CC-BY-NC-SA-4.0 | test | curated:b137b549e3902ec4 |
0191563e290b6194e678b820f5de0da6 | 0191563e290b6194e678b820f5de0da6 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Read the contents of a file at the given path. Returns the file content as a string.","name":"read_file","parameters":{"properties":{"path":{"description":"Absolute or relative file path","type":"string"}},"required":["path"],"type":"object"}},"type":"function"},{"function":{"description":"... | code_debug | glm5.2-agent-tool-synthetic | agent_tool | code_debug | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"code_debug","expected_final":"The deterministic test must pass after correcting /tmp/agent_tool_workspace_11000457/src/windows.py so that window([2, 4, 6]) returns 30; report the test output and the corrected behavior.","expected_tools":["read_file","patch_file","run_python"],"seed":11000457} | sha256:446331c0b65bf1faef205c4c59bf7f8c7a73a8f0f6a3b2d9d4691957c71a9e29 | 7bb0d53266441f6ecb548cd2 | 9d7ee60a60fe54ae2ea27873 | f6e5c42a85b34839f10a70ed | true | true | null | null | [] | 1.174049 | 4 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-code_debug-11000457 |
0195f9152467fb461149cf3bcffd7747 | 0195f9152467fb461149cf3bcffd7747 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | Dolly | Dolly | instruction | Dolly | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Gisborne City" | sha256:b69c87ad3cca84a8ec4e1abaf790c13435118252c5359b810df455375e64af40 | 1b0d665f158a6411ab8a04ed | 944db36053f565f7351b0489 | d43d4880c2d51c054cdb8bef | null | null | null | 9 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | databricks/databricks-dolly-15k | bdd27f4d94b9c1f951818a7da7fd7aeea5dbff1a | CC-BY-SA-3.0 | test | curated:34bd1161bc2b9e87 |
019a602d6a95554612c825bb56f858ec | 019a602d6a95554612c825bb56f858ec | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:e228518910e511ea344318b38e70910f3133436ecb9b254b0c144c49af553572 | 61e3ea30675c10d75c82345c | c55b423282a9160ec75cb07d | d8402cc643e3bcda15e74aea | null | null | null | 6 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:04e0d180e71cef1b |
019a8b44acac877d0e3cde5572e7e615 | 019a8b44acac877d0e3cde5572e7e615 | train | [{"content":"You are an expert logical reasoner. Think through the problem carefully.\n\nFormat your response EXACTLY as:\n<think>\n[Break down the problem into components.\nConsider multiple hypotheses if applicable.\nEliminate incorrect paths explicitly.\nBuild toward the conclusion step by step.]\n</think>\n\n[Clear... | [] | SciQ | SciQ | reasoning | SciQ | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "B" | sha256:2a6b3ebea2df8a4249c524bc2a01bf375726c3f2faaa16d9ce038b42ebc6af23 | 45dce950b7c723a1a470925b | 2b411f44ec2ef96614ee5aa5 | 758d496f3b90cd7c5b2f8bd7 | true | null | true | 9.5 | [] | 1 | 1 | 0 | 1.0.0 | allenai/sciq | 2c94ad3e1aafab77146f384e23536f97a4849815 | CC-BY-NC-3.0 | train | curated:c547cfbeb8f1e69a |
019b2ec93200030b06690053708ee585 | 019b2ec93200030b06690053708ee585 | train | [{"content":"You are a helpful, precise assistant. Follow instructions exactly.\n\nFormat your response as:\n<think>\n[Understand what is being asked. Identify constraints and requirements.\nPlan your response structure.]\n</think>\n\n[Your complete, well-structured response following all given constraints]","name":"",... | [] | tulu-3 | tulu-3 | instruction | tulu-3 | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "" | sha256:3b887b999f98bd5a4e5d0a94cef53365da26c224923cded1c66bb1006dbbf608 | 100ee05dcd1e465f924138e8 | 417943fafb7983c4d7c85537 | aacc086deceb3ee6127cb46e | null | null | null | 7.5 | [
"reference_status_unknown",
"reference_unknown"
] | 1 | 1 | 0 | 1.0.0 | allenai/tulu-3-sft-mixture | b14afda60f1bbebe55d5d2fa1e4df5042f97f8be | ODC-BY-1.0; mixture terms vary | train | curated:6a909036adc3317f |
019b8a2fd50eead7184573c36aa5343d | 019b8a2fd50eead7184573c36aa5343d | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search for files matching a pattern or containing a text pattern.","name":"search_files","parameters":{"properties":{"path":{"default":".","description":"Directory to search in","type":"string"},"pattern":{"description":"Search pattern (glob or text)","type":"string"}},"required":["pattern"... | file_ops | glm5.2-agent-tool-synthetic | agent_tool | file_ops | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"file_ops","expected_final":"Search for the marker, verify the source, replace 'RETRY_LIMIT = 4' with 'RETRY_LIMIT = 5' in /tmp/agent_tool_workspace_10000832/src/worker.py, and read the file back to confirm the exact edit.","expected_tools":["search_files","read_file","patch_file","read_file"],"seed":100008... | sha256:b90934bbd675e5ccb10a37d292f2a273fb5c9424b7074cfe907d01d96cc7a06f | 8e89c399bc9a7ca5bacc2f42 | 1f14369079fcfe1cbb24c6d0 | bd351edb6bdf66da4d4537c9 | true | true | null | null | [] | 1.611823 | 4 | 4 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-file_ops-10000832 |
019c477eb90ea058256ac8ea6b807808 | 019c477eb90ea058256ac8ea6b807808 | train | [{"content":"You are a careful deterministic workspace assistant. The workspace and local corpus described by the user are synthetic ephemeral fixtures, not the user's live files or live services. Solve the task by using the provided tools through native OpenAI function calling (the assistant message must contain a too... | [{"function":{"description":"Search a fixed local document corpus and return matching passages.","name":"corpus_search","parameters":{"properties":{"query":{"description":"Search query","type":"string"},"top_k":{"default":5,"description":"Number of results to return","type":"integer"}},"required":["query"],"type":"obje... | multi_research | glm5.2-agent-tool-synthetic | agent_tool | multi_research | glm-5.2 | z.ai-coding-paas-v4 | multi_turn_tool | traj_success | {"category":"multi_research","expected_final":"Cite the retrieved passages for batch size=24 and latency=320, then report the calculator result for 320 / 24 as 13.3333.","expected_tools":["corpus_search","corpus_search","calculator"],"seed":13001279} | sha256:2f56d5a24021b60d464a5a2b815ba036bf83fecd81456a3e959986579f9a5ab2 | bcd87b1ad319c2706dbeae6c | b351ad1a80972e0bf0a6351d | fc6250b1fcb579392960d075 | true | true | null | null | [] | 1.611823 | 3 | 3 | 1.0.0 | local:distillation-glm52 | run:full-20260724-175443@sha256:71f34bba1cb29a831fd1f26b51f0b787d9d4ad2bc51c93cb1af5084759e128e1 | other; synthetic research corpus; see provenance | train | agent-multi_research-13001279 |
019e29789d0b2f14ecc76bb4791fad80 | 019e29789d0b2f14ecc76bb4791fad80 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | OrcaMath | OrcaMath | math | OrcaMath | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "To find out how many two-digit numbers divisible by 3 can be made using the cards 1, 2, and 6, we need to consider all possible combinations of these two digits and then check which of them are divisible by 3.\n\nThe possible two-digit numbers we can make are:\n- 12\n- 16\n- 21\n- 26\n- 61\n- 62\n\nNow, let's check wh... | sha256:ae782b7a341be697c9f2a6811c2b782675956452c421b97363baf2302df5003b | 55b651d4554afb5a47053922 | 469db5fe6081d8916758651b | 8dbeb6c7cc78b33e5d592849 | true | null | true | 9 | [] | 1 | 1 | 0 | 1.0.0 | microsoft/orca-math-word-problems-200k | 29255d1770cc4eac66e5e7fa378cba542c026350 | unknown-see-provenance | train | curated:66020d52d13a7743 |
01a05b45e26c339c9d9a9e121e25cd14 | 01a05b45e26c339c9d9a9e121e25cd14 | train | [{"content":"You are an expert mathematician. Solve the given problem with complete rigor.\n\nFormat your response EXACTLY as:\n<think>\n[Your complete step-by-step reasoning. Show all intermediate calculations.\nIf you make an error and catch it, show the correction.\nEnd with a clear statement of the final answer.]\n... | [] | MetaMathQA | MetaMathQA | math | MetaMathQA | qwen3.8-max-preview | alibaba-model-studio | single_turn | sft | "Tracy received 300 book donations.\nTen people came and donated five books each, so they donated a total of 10 x 5 = 50 books.\nThe total number of books in the store is now 300 + 50 = 350 books.\nA total of 140 books were borrowed by other people to read.\nTherefore, Tracy has 350 - 140 = 210 books remaining in the s... | sha256:946664e845bc5eb072b9ce3a50acea10231999b48c9d82adf5488b267334efa0 | 02c8654bfa786218df70f481 | 10a9219429a8669d29e6aeef | 91a1c65c3fdfe132d0557201 | true | null | true | 10 | [
"difficulty_untrusted",
"ground_truth_restored_from_pinned_upstream"
] | 1 | 1 | 0 | 1.0.0 | meta-math/MetaMathQA | aa4f34d3d2d3231299b5b03d9b3e5a20da45aa18 | MIT | train | curated:09897cf966a6bfcc |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.