code-adapter-harness-evidence / plans /question_first.jsonl
Corning's picture
Publish selected code-adapter trajectories and harness evidence
4d3c068 verified
Raw History Blame Contribute Delete
43.2 kB
{"task_id": "scibench_atkins_065", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_atkins_042", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_atkins_015", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_calculus_023", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_calculus_036", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_calculus_037", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_chemmc_018", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_chemmc_006", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_chemmc_002", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_class_049", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_class_051", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_class_010", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_diff_021", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_diff_010", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_diff_032", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_fund_017", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_fund_004", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_fund_055", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_matter_015", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_matter_023", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_matter_021", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_quan_013", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_quan_032", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_quan_003", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_stat_003", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_stat_058", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_stat_016", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_thermo_014", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_thermo_055", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}
{"task_id": "scibench_thermo_054", "plan_text": "Derive a fresh solution directly from the original problem above. No prior solution or candidate value is supplied.\nFirst identify the exact final requested quantity and its required answer unit. For a multipart problem that requires one scalar submission, compute the preceding parts as intermediate quantities and submit the final requested quantity. A blank unit does not specify a physical unit; infer dimensions from the requested quantity.\nBuild the model from the stated assumptions, initial conditions, and constraints. Distinguish a change from an absolute final value. Account for stated physical effects unless the question explicitly neglects them. Convert units consistently before calculation.\nCompute with Python. Test the result against an independent implication of the original question, such as a stated limiting value, initial condition, conservation law, dimensions, or an independent numerical formulation. Substitution into the same assumed formula checks arithmetic, not whether that formula describes the problem. If a check contradicts the original problem, repair the model before submitting.\nPrint the requested quantity and its unit separately from intermediate values. Finish once the calculation and a meaningful check support it; do not repeatedly reprint a number. Use ordinary text and the existing sandbox tools, without any mandatory report schema."}