Dataset Viewer
Auto-converted to Parquet
data_source
stringclasses
1 value
prompt
listlengths
1
1
ability
stringclasses
1 value
reward_model
dict
extra_info
dict
id
stringlengths
7
10
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t P \t X \t _ \t O \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 0, "split": "train" }
train_0
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t P \t X \t O \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 1, "split": "train" }
train_1
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t O \t X \t _ \t # \t\n # \t _ \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 2, "split": "train" }
train_2
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t O \t X \t _ \t # \t\n # \t _ \t _ \t _ \t P \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 3, "split": "train" }
train_3
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t O \t X \t P \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 4, "split": "train" }
train_4
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t _ \t X \t _ \t # \t\n # \t # \t # \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 5, "split": "train" }
train_5
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t O \t _ \t X \t _ \t # \t\n # \t # \t _ \t P \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 6, "split": "train" }
train_6
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t O \t _ \t X \t _ \t # \t\n # \t # \t _ \t _ \t P \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 7, "split": "train" }
train_7
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t O \t _ \t X \t P \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 8, "split": "train" }
train_8
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t O \t X \t P \t _ \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 9, "split": "train" }
train_9
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 10, "split": "train" }
train_10
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 11, "split": "train" }
train_11
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t _ \t _ \t P \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 12, "split": "train" }
train_12
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t _ \t _ \t X \t P \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 13, "split": "train" }
train_13
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t _ \t X \t P \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 14, "split": "train" }
train_14
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t X \t P \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 15, "split": "train" }
train_15
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 16, "split": "train" }
train_16
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t # \t # \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t P \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 17, "split": "train" }
train_17
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t _ \t O \t # \t # \t\n # \t _ \t # \t X \t _ \t # \t\n # \t _ \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 18, "split": "train" }
train_18
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t P \t X \t O \t _ \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 19, "split": "train" }
train_19
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t _ \t X \t _ \t # \t\n # \t # \t P \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 20, "split": "train" }
train_20
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t P \t X \t _ \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 21, "split": "train" }
train_21
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t _ \t P \t X \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 22, "split": "train" }
train_22
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t _ \t _ \t X \t # \t\n # \t # \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 23, "split": "train" }
train_23
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t _ \t _ \t X \t # \t\n # \t # \t _ \t _ \t P \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 24, "split": "train" }
train_24
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t _ \t X \t _ \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 25, "split": "train" }
train_25
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t P \t _ \t _ \t _ \t # \t\n # \t _ \t X \t _ \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 26, "split": "train" }
train_26
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t P \t X \t _ \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 27, "split": "train" }
train_27
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t P \t X \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 28, "split": "train" }
train_28
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 29, "split": "train" }
train_29
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t P \t _ \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 30, "split": "train" }
train_30
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t O \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t X \t # \t # \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 31, "split": "train" }
train_31
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t O \t _ \t _ \t # \t\n # \t _ \t X \t _ \t _ \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 32, "split": "train" }
train_32
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t X \t P \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 33, "split": "train" }
train_33
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t X \t P \t _ \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 34, "split": "train" }
train_34
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t X \t P \t _ \t _ \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 35, "split": "train" }
train_35
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 36, "split": "train" }
train_36
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t P \t _ \t # \t # \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 37, "split": "train" }
train_37
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t P \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 38, "split": "train" }
train_38
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t X \t P \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 39, "split": "train" }
train_39
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t X \t P \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 40, "split": "train" }
train_40
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t X \t P \t _ \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 41, "split": "train" }
train_41
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 42, "split": "train" }
train_42
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t P \t _ \t _ \t _ \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 43, "split": "train" }
train_43
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t P \t # \t\n # \t # \t # \t _ \t X \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t _ \t O \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 44, "split": "train" }
train_44
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t _ \t P \t # \t\n # \t # \t # \t # \t X \t # \t\n # \t # \t # \t _ \t O \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 45, "split": "train" }
train_45
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t X \t O \t _ \t # \t\n # \t P \t # \t _ \t _ \t # \t\n # \t _ \t # \t # \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 46, "split": "train" }
train_46
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t P \t X \t O \t _ \t # \t\n # \t _ \t # \t _ \t _ \t # \t\n # \t _ \t # \t # \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 47, "split": "train" }
train_47
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t _ \t X \t # \t # \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 48, "split": "train" }
train_48
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t X \t _ \t O \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 49, "split": "train" }
train_49
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t X \t _ \t O \t # \t\n # \t P \t _ \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 50, "split": "train" }
train_50
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t P \t X \t _ \t O \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 51, "split": "train" }
train_51
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t P \t X \t O \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 52, "split": "train" }
train_52
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t P \t X \t _ \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 53, "split": "train" }
train_53
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t _ \t P \t X \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 54, "split": "train" }
train_54
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t _ \t _ \t P \t X \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 55, "split": "train" }
train_55
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t _ \t _ \t _ \t X \t # \t\n # \t # \t # \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 56, "split": "train" }
train_56
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t _ \t _ \t _ \t X \t # \t\n # \t # \t # \t _ \t P \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 57, "split": "train" }
train_57
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t O \t # \t\n # \t _ \t X \t _ \t _ \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 58, "split": "train" }
train_58
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t X \t _ \t O \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 59, "split": "train" }
train_59
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t X \t _ \t O \t # \t\n # \t P \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 60, "split": "train" }
train_60
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t P \t X \t _ \t O \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 61, "split": "train" }
train_61
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t P \t X \t O \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 62, "split": "train" }
train_62
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t X \t _ \t # \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 63, "split": "train" }
train_63
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t X \t _ \t _ \t # \t\n # \t _ \t P \t _ \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 64, "split": "train" }
train_64
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t X \t _ \t _ \t # \t\n # \t _ \t _ \t P \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 65, "split": "train" }
train_65
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t X \t P \t _ \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 66, "split": "train" }
train_66
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t P \t X \t _ \t _ \t # \t\n # \t _ \t _ \t O \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 67, "split": "train" }
train_67
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t _ \t P \t X \t _ \t # \t\n # \t _ \t _ \t O \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 68, "split": "train" }
train_68
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t P \t O \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 69, "split": "train" }
train_69
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t _ \t S \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 70, "split": "train" }
train_70
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t _ \t O \t P \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 71, "split": "train" }
train_71
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t _ \t _ \t X \t P \t # \t\n # \t _ \t _ \t O \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 72, "split": "train" }
train_72
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t P \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t _ \t O \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 73, "split": "train" }
train_73
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t P \t _ \t # \t\n # \t _ \t _ \t X \t _ \t # \t\n # \t _ \t _ \t O \t _ \t # \t\n # \t # \t # \t # \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 74, "split": "train" }
train_74
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t X \t P \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 75, "split": "train" }
train_75
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t X \t P \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 76, "split": "train" }
train_76
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t _ \t # \t X \t _ \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 77, "split": "train" }
train_77
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t O \t _ \t # \t\n # \t _ \t # \t X \t _ \t # \t\n # \t _ \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 78, "split": "train" }
train_78
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t _ \t X \t P \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 79, "split": "train" }
train_79
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t O \t X \t P \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 80, "split": "train" }
train_80
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t _ \t _ \t # \t # \t\n # \t P \t X \t _ \t O \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 81, "split": "train" }
train_81
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t # \t # \t # \t\n # \t # \t _ \t _ \t # \t # \t\n # \t _ \t P \t X \t O \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 82, "split": "train" }
train_82
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t _ \t P \t _ \t # \t\n # \t # \t _ \t X \t _ \t # \t\n # \t # \t _ \t O \t _ \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 83, "split": "train" }
train_83
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t O \t _ \t _ \t # \t\n # \t _ \t X \t _ \t # \t # \t\n # \t P \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 84, "split": "train" }
train_84
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t _ \t O \t _ \t _ \t # \t\n # \t _ \t X \t _ \t # \t # \t\n # \t _ \t P \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 85, "split": "train" }
train_85
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t _ \t O \t # \t\n # \t # \t # \t # \t X \t # \t\n # \t # \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 86, "split": "train" }
train_86
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t # \t _ \t _ \t # \t\n # \t # \t # \t _ \t O \t # \t\n # \t # \t # \t # \t X \t # \t\n # \t # \t _ \t _ \t P \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 87, "split": "train" }
train_87
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t _ \t X \t P \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 88, "split": "train" }
train_88
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t X \t P \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 89, "split": "train" }
train_89
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t X \t P \t _ \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 90, "split": "train" }
train_90
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 91, "split": "train" }
train_91
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t P \t _ \t # \t # \t # \t\n # \t X \t _ \t _ \t _ \t # \t\n # \t O \t _ \t _ \t _ \t # \t\n # \t _ \t _ \t _ \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 92, "split": "train" }
train_92
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t # \t _ \t X \t O \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 93, "split": "train" }
train_93
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t # \t P \t X \t O \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 4 ], "style": "rule" }
{ "index": 94, "split": "train" }
train_94
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t P \t # \t # \t\n # \t _ \t X \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t O \t _ \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 95, "split": "train" }
train_95
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t P \t _ \t # \t # \t\n # \t _ \t X \t # \t # \t # \t\n # \t _ \t _ \t # \t # \t # \t\n # \t _ \t O \t _ \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 96, "split": "train" }
train_96
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t _ \t _ \t _ \t # \t # \t\n # \t _ \t P \t # \t # \t # \t\n # \t _ \t X \t # \t # \t # \t\n # \t _ \t O \t _ \t # \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 2 ], "style": "rule" }
{ "index": 97, "split": "train" }
train_97
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t _ \t X \t _ \t # \t\n # \t # \t _ \t _ \t P \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 3 ], "style": "rule" }
{ "index": 98, "split": "train" }
train_98
sokoban_6x6_1horizon
[ { "content": "<|im_start|>user\nYou are a Sokoban solver.\n\nSokoban Quick Guide\nGoal: Push all boxes (X) onto targets (O).\n\nSymbols:\n# Wall | _ Floor | O Target | X Box | P You | √ = Box on Target | S = You on Target\n\nRules:\n1. Push boxes (can't pull).\n2. Avoid walls (#).\n\nAnswers:\n<answer> Up </answer> | <answer> Down </answer> | <answer> Left </answer> | <answer> Right </answer>\n\nRewards:\nMove: -0.1\nBox on target: +1.0\nAll boxes placed: +10.0\n\n\n[Current Observation]:\n # \t # \t # \t # \t # \t # \t\n # \t # \t _ \t _ \t O \t # \t\n # \t # \t _ \t _ \t _ \t # \t\n # \t # \t _ \t X \t _ \t # \t\n # \t # \t _ \t P \t _ \t # \t\n # \t # \t # \t # \t # \t # \t\nDecide the next action:\nAlways output: <think> [Your thoughts] </think> <answer> [your answer] </answer> with no extra text. Strictly follow this format. <|im_end|>\n<|im_start|>assistant\n<think>", "role": "user" } ]
bfs
{ "ground_truth": [ 1 ], "style": "rule" }
{ "index": 99, "split": "train" }
train_99
End of preview. Expand in Data Studio
YAML Metadata Warning: The task_categories "TEXT_GENERATION" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

Debunk the Myth of SFT Generalization Dataset

This dataset is associated with the paper Debunk the Myth of SFT Generalization.

The research presented in the paper challenges the common belief that supervised fine-tuning (SFT) primarily memorizes training data and lacks generalization, especially when compared to reinforcement learning (RL). Through systematic evaluation on decision-making benchmarks like Sokoban and General Points, the authors demonstrate that incorporating prompt diversity and Chain-of-Thought (CoT) supervision during SFT training can lead to strong generalization across unseen instruction variants and strictly harder tasks, often matching or surpassing RL baselines.

Code: https://github.com/XiaofengLin7/debunking-sft-generalization

Installation

This dataset is intended for use with the associated code repository. To set up the environment and dependencies:

Prerequisites

CUDA 12.2 & cuDNN 9.1.0 works, but official docs recommends CUDA >= 12.4 & cuDNN >= 9.8.0.

Setup

conda create -n debunk_sft python=3.10
conda activate debunk_sft
USE_MEGATRON=0 bash setup.sh
git submodule init
git submodule update
pip install -e thirdparty/verl --no-dependencies
pip install -e thirdparty/ragen --no-dependencies
pip install -e thirdparty/alfworld --no-dependencies
pip install -e thirdparty/trl --no-dependecies

Dataset Overview

This repository contains various datasets used in the research, categorized by task, method, diversity, and format. These datasets are part of a larger collection.

Task Method Diversity Format Link
Sokoban RL non-diverse β€” πŸ€—
Sokoban RL diverse β€” πŸ€—
Sokoban SFT non-diverse answer-only πŸ€—
Sokoban SFT diverse answer-only πŸ€—
Sokoban SFT non-diverse cot πŸ€—
Sokoban SFT diverse cot πŸ€—
General Points RL non-diverse β€” πŸ€—
General Points RL diverse β€” πŸ€—
General Points SFT non-diverse answer-only πŸ€—
General Points SFT diverse answer-only πŸ€—
General Points SFT non-diverse cot πŸ€—
General Points SFT diverse cot πŸ€—

Sample Usage

The following snippets from the GitHub repository demonstrate how to train models using this dataset with SFT (Supervised Fine-Tuning) or GRPO (Generative Reinforcement Policy Optimization).

Train your model with SFT

Specify your model and data beforehand.

For Sokoban:

bash debunk_sft/scripts/sokoban/sokoban_train_and_eval.sh

For General Points:

bash debunk_sft/scripts/gp_l/gp_l_train_and_eval.sh

Train your model with GRPO

Specify your model and data beforehand.

For Sokoban:

bash debunk_sft/scripts/sokoban/sokoban_grpo.sh

For General Points:

bash debunk_sft/scripts/gp_l/gp_l_grpo.sh

Citation

If you use this dataset in your research, please cite the associated paper:

@misc{lin2024debunkthemythofsftgeneralization,
      title={Debunk the Myth of SFT Generalization},
      author={Xiaofeng Lin and Yuandong Tian and Huazhe Xu},
      year={2024},
      eprint={2510.00237},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2510.00237},
}
Downloads last month
25

Collection including Xiaofeng77/sokoban