prompt listlengths 2 2 | label dict | metadata dict |
|---|---|---|
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{R^{2}\\sqrt{2}}{4}",
"budget": 6330
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003185",
"normalized_example_id": "ex_003185",
"source_idx": 3185,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6330,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "57",
"budget": 1520
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003186",
"normalized_example_id": "ex_003186",
"source_idx": 3186,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2534,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "57",
"budget": 2027
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003186",
"normalized_example_id": "ex_003186",
"source_idx": 3186,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2534,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "57",
"budget": 2534
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003186",
"normalized_example_id": "ex_003186",
"source_idx": 3186,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2534,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "1.6461",
"budget": 4384
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003187",
"normalized_example_id": "ex_003187",
"source_idx": 3187,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7306,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "1.6461",
"budget": 5845
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003187",
"normalized_example_id": "ex_003187",
"source_idx": 3187,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7306,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "1.6461",
"budget": 7306
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003187",
"normalized_example_id": "ex_003187",
"source_idx": 3187,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7306,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "a \\geqslant-\\frac{53}{28}",
"budget": 3478
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003189",
"normalized_example_id": "ex_003189",
"source_idx": 3189,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5796,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "a \\geqslant-\\frac{53}{28}",
"budget": 4637
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003189",
"normalized_example_id": "ex_003189",
"source_idx": 3189,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5796,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "a \\geqslant-\\frac{53}{28}",
"budget": 5796
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003189",
"normalized_example_id": "ex_003189",
"source_idx": 3189,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5796,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "24",
"budget": 949
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003190",
"normalized_example_id": "ex_003190",
"source_idx": 3190,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 1582,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "24",
"budget": 1266
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003190",
"normalized_example_id": "ex_003190",
"source_idx": 3190,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 1582,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "24",
"budget": 1582
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003190",
"normalized_example_id": "ex_003190",
"source_idx": 3190,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 1582,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "96",
"budget": 1721
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003191",
"normalized_example_id": "ex_003191",
"source_idx": 3191,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2869,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "96",
"budget": 2295
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003191",
"normalized_example_id": "ex_003191",
"source_idx": 3191,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2869,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "96",
"budget": 2869
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003191",
"normalized_example_id": "ex_003191",
"source_idx": 3191,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2869,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "500000",
"budget": 4257
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003192",
"normalized_example_id": "ex_003192",
"source_idx": 3192,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7095,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "500000",
"budget": 5676
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003192",
"normalized_example_id": "ex_003192",
"source_idx": 3192,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7095,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "500000",
"budget": 7095
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003192",
"normalized_example_id": "ex_003192",
"source_idx": 3192,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7095,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{\\sqrt{86}}{86}",
"budget": 3293
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003194",
"normalized_example_id": "ex_003194",
"source_idx": 3194,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5489,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{\\sqrt{86}}{86}",
"budget": 4391
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003194",
"normalized_example_id": "ex_003194",
"source_idx": 3194,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5489,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{\\sqrt{86}}{86}",
"budget": 5489
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003194",
"normalized_example_id": "ex_003194",
"source_idx": 3194,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5489,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "32\\pi",
"budget": 2812
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003195",
"normalized_example_id": "ex_003195",
"source_idx": 3195,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4687,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "32\\pi",
"budget": 3750
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003195",
"normalized_example_id": "ex_003195",
"source_idx": 3195,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4687,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "32\\pi",
"budget": 4687
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003195",
"normalized_example_id": "ex_003195",
"source_idx": 3195,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4687,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "75^\\circ",
"budget": 4382
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003196",
"normalized_example_id": "ex_003196",
"source_idx": 3196,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7303,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "75^\\circ",
"budget": 5842
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003196",
"normalized_example_id": "ex_003196",
"source_idx": 3196,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7303,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "75^\\circ",
"budget": 7303
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003196",
"normalized_example_id": "ex_003196",
"source_idx": 3196,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7303,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "(x,y)\\in{(0,0),(\\frac{1}{2}(1+\\sqrt{5}),\\frac{1}{2}(1+\\sqrt{5})),(\\frac{1}{2}(1-\\sqrt{5}),\\frac{1}{2}(1-\\sqrt{5}))}",
"budget": 3577
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003197",
"normalized_example_id": "ex_003197",
"source_idx": 3197,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5962,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "(x,y)\\in{(0,0),(\\frac{1}{2}(1+\\sqrt{5}),\\frac{1}{2}(1+\\sqrt{5})),(\\frac{1}{2}(1-\\sqrt{5}),\\frac{1}{2}(1-\\sqrt{5}))}",
"budget": 4770
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003197",
"normalized_example_id": "ex_003197",
"source_idx": 3197,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5962,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "(x,y)\\in{(0,0),(\\frac{1}{2}(1+\\sqrt{5}),\\frac{1}{2}(1+\\sqrt{5})),(\\frac{1}{2}(1-\\sqrt{5}),\\frac{1}{2}(1-\\sqrt{5}))}",
"budget": 5962
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003197",
"normalized_example_id": "ex_003197",
"source_idx": 3197,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5962,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2,1",
"budget": 2479
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003198",
"normalized_example_id": "ex_003198",
"source_idx": 3198,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4132,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2,1",
"budget": 3306
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003198",
"normalized_example_id": "ex_003198",
"source_idx": 3198,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4132,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2,1",
"budget": 4132
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003198",
"normalized_example_id": "ex_003198",
"source_idx": 3198,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4132,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "(3,4,7)",
"budget": 2959
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003199",
"normalized_example_id": "ex_003199",
"source_idx": 3199,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4931,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "(3,4,7)",
"budget": 3945
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003199",
"normalized_example_id": "ex_003199",
"source_idx": 3199,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4931,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "(3,4,7)",
"budget": 4931
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003199",
"normalized_example_id": "ex_003199",
"source_idx": 3199,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4931,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "26",
"budget": 2124
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003201",
"normalized_example_id": "ex_003201",
"source_idx": 3201,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3540,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "26",
"budget": 2832
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003201",
"normalized_example_id": "ex_003201",
"source_idx": 3201,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3540,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "26",
"budget": 3540
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003201",
"normalized_example_id": "ex_003201",
"source_idx": 3201,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3540,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "3^{400}>4^{300}",
"budget": 4258
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003202",
"normalized_example_id": "ex_003202",
"source_idx": 3202,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7097,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "3^{400}>4^{300}",
"budget": 5678
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003202",
"normalized_example_id": "ex_003202",
"source_idx": 3202,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7097,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "3^{400}>4^{300}",
"budget": 7097
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003202",
"normalized_example_id": "ex_003202",
"source_idx": 3202,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7097,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "28\\cdot1+30\\cdot4+31\\cdot7=365",
"budget": 3117
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003203",
"normalized_example_id": "ex_003203",
"source_idx": 3203,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5195,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "28\\cdot1+30\\cdot4+31\\cdot7=365",
"budget": 4156
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003203",
"normalized_example_id": "ex_003203",
"source_idx": 3203,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5195,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "28\\cdot1+30\\cdot4+31\\cdot7=365",
"budget": 5195
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003203",
"normalized_example_id": "ex_003203",
"source_idx": 3203,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5195,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "26",
"budget": 1221
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003204",
"normalized_example_id": "ex_003204",
"source_idx": 3204,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2035,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "26",
"budget": 1628
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003204",
"normalized_example_id": "ex_003204",
"source_idx": 3204,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2035,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "26",
"budget": 2035
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003204",
"normalized_example_id": "ex_003204",
"source_idx": 3204,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2035,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "45",
"budget": 1260
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003205",
"normalized_example_id": "ex_003205",
"source_idx": 3205,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2100,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "45",
"budget": 1680
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003205",
"normalized_example_id": "ex_003205",
"source_idx": 3205,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2100,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "45",
"budget": 2100
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003205",
"normalized_example_id": "ex_003205",
"source_idx": 3205,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2100,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "27100",
"budget": 5896
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003206",
"normalized_example_id": "ex_003206",
"source_idx": 3206,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 9826,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "27100",
"budget": 7861
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003206",
"normalized_example_id": "ex_003206",
"source_idx": 3206,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 9826,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "27100",
"budget": 9826
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003206",
"normalized_example_id": "ex_003206",
"source_idx": 3206,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 9826,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "-2015",
"budget": 2220
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003212",
"normalized_example_id": "ex_003212",
"source_idx": 3212,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3700,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "-2015",
"budget": 2960
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003212",
"normalized_example_id": "ex_003212",
"source_idx": 3212,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3700,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "-2015",
"budget": 3700
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003212",
"normalized_example_id": "ex_003212",
"source_idx": 3212,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3700,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "0.63",
"budget": 2929
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003213",
"normalized_example_id": "ex_003213",
"source_idx": 3213,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4882,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "0.63",
"budget": 3906
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003213",
"normalized_example_id": "ex_003213",
"source_idx": 3213,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4882,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "0.63",
"budget": 4882
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003213",
"normalized_example_id": "ex_003213",
"source_idx": 3213,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 4882,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "30",
"budget": 3841
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003214",
"normalized_example_id": "ex_003214",
"source_idx": 3214,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6401,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "30",
"budget": 5121
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003214",
"normalized_example_id": "ex_003214",
"source_idx": 3214,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6401,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "30",
"budget": 6401
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003214",
"normalized_example_id": "ex_003214",
"source_idx": 3214,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6401,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "14,15,16",
"budget": 3374
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003215",
"normalized_example_id": "ex_003215",
"source_idx": 3215,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5624,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "14,15,16",
"budget": 4499
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003215",
"normalized_example_id": "ex_003215",
"source_idx": 3215,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5624,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "14,15,16",
"budget": 5624
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003215",
"normalized_example_id": "ex_003215",
"source_idx": 3215,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5624,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2002",
"budget": 4154
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003216",
"normalized_example_id": "ex_003216",
"source_idx": 3216,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6923,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2002",
"budget": 5538
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003216",
"normalized_example_id": "ex_003216",
"source_idx": 3216,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6923,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2002",
"budget": 6923
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003216",
"normalized_example_id": "ex_003216",
"source_idx": 3216,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 6923,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "3a-2b",
"budget": 1385
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003221",
"normalized_example_id": "ex_003221",
"source_idx": 3221,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2308,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "3a-2b",
"budget": 1846
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003221",
"normalized_example_id": "ex_003221",
"source_idx": 3221,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2308,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "3a-2b",
"budget": 2308
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003221",
"normalized_example_id": "ex_003221",
"source_idx": 3221,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2308,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "9",
"budget": 1628
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003223",
"normalized_example_id": "ex_003223",
"source_idx": 3223,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2714,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "9",
"budget": 2171
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003223",
"normalized_example_id": "ex_003223",
"source_idx": 3223,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2714,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "9",
"budget": 2714
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003223",
"normalized_example_id": "ex_003223",
"source_idx": 3223,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2714,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "C",
"budget": 5993
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003225",
"normalized_example_id": "ex_003225",
"source_idx": 3225,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 9988,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "C",
"budget": 7990
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003225",
"normalized_example_id": "ex_003225",
"source_idx": 3225,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 9988,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "C",
"budget": 9988
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003225",
"normalized_example_id": "ex_003225",
"source_idx": 3225,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 9988,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2\\sqrt{2}",
"budget": 3458
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003226",
"normalized_example_id": "ex_003226",
"source_idx": 3226,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5764,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2\\sqrt{2}",
"budget": 4611
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003226",
"normalized_example_id": "ex_003226",
"source_idx": 3226,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5764,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "2\\sqrt{2}",
"budget": 5764
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003226",
"normalized_example_id": "ex_003226",
"source_idx": 3226,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5764,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "91,49,28",
"budget": 2199
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003231",
"normalized_example_id": "ex_003231",
"source_idx": 3231,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3665,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "91,49,28",
"budget": 2932
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003231",
"normalized_example_id": "ex_003231",
"source_idx": 3231,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3665,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "91,49,28",
"budget": 3665
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003231",
"normalized_example_id": "ex_003231",
"source_idx": 3231,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 3665,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\sqrt{5}-1",
"budget": 4396
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003232",
"normalized_example_id": "ex_003232",
"source_idx": 3232,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7326,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\sqrt{5}-1",
"budget": 5861
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003232",
"normalized_example_id": "ex_003232",
"source_idx": 3232,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7326,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\sqrt{5}-1",
"budget": 7326
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003232",
"normalized_example_id": "ex_003232",
"source_idx": 3232,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7326,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "20",
"budget": 3248
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003235",
"normalized_example_id": "ex_003235",
"source_idx": 3235,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5414,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "20",
"budget": 4331
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003235",
"normalized_example_id": "ex_003235",
"source_idx": 3235,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5414,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "20",
"budget": 5414
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003235",
"normalized_example_id": "ex_003235",
"source_idx": 3235,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 5414,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "-\\frac{8}{7}(\\sqrt[4]{\\frac{1+\\sqrt{x}}{\\sqrt{x}}})^{7}+C",
"budget": 4229
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003237",
"normalized_example_id": "ex_003237",
"source_idx": 3237,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7049,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "-\\frac{8}{7}(\\sqrt[4]{\\frac{1+\\sqrt{x}}{\\sqrt{x}}})^{7}+C",
"budget": 5639
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003237",
"normalized_example_id": "ex_003237",
"source_idx": 3237,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7049,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "-\\frac{8}{7}(\\sqrt[4]{\\frac{1+\\sqrt{x}}{\\sqrt{x}}})^{7}+C",
"budget": 7049
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003237",
"normalized_example_id": "ex_003237",
"source_idx": 3237,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 7049,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "21",
"budget": 1070
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003242",
"normalized_example_id": "ex_003242",
"source_idx": 3242,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 1784,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "21",
"budget": 1427
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003242",
"normalized_example_id": "ex_003242",
"source_idx": 3242,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 1784,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "21",
"budget": 1784
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003242",
"normalized_example_id": "ex_003242",
"source_idx": 3242,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 1784,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{1}{2} \\tan \\theta",
"budget": 1525
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003247",
"normalized_example_id": "ex_003247",
"source_idx": 3247,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2542,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{1}{2} \\tan \\theta",
"budget": 2034
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003247",
"normalized_example_id": "ex_003247",
"source_idx": 3247,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2542,
... |
[
{
"role": "system",
"content": "You are a reasoning control agent that guides a step-by-step mathematical reasoner.\n\nEach turn you receive the reasoner's latest step and the remaining token budget (as a percentage). On the first turn you receive the question instead. Considering the reasoner's current pro... | {
"answer": "\\frac{1}{2} \\tan \\theta",
"budget": 2542
} | {
"data_type": "full_prompt_per_prompt_calibrated_budget",
"source_example_id": "ex_003247",
"normalized_example_id": "ex_003247",
"source_idx": 3247,
"augmentation": "online_rl_full_prompt_per_prompt_calibrated_budget",
"budget_source": "per_prompt_calibrated_length_ratios",
"calibrated_length": 2542,
... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.