Dataset Viewer
Auto-converted to Parquet Duplicate
question
stringlengths
135
141
metadata
stringlengths
160
172
task_source
stringclasses
1 value
formatted_prompt
listlengths
2
2
responses_by_sample
listlengths
1
1
tokens_by_sample
listlengths
1
1
num_rounds_by_sample
listlengths
1
1
continuation_prompts_by_sample
listlengths
1
1
continuation_text_by_sample
listlengths
1
1
strategy_type
stringclasses
1 value
termination_reason_by_sample
listlengths
1
1
correct_round_by_sample
listlengths
1
1
smart_termination_enabled
bool
1 class
Using the numbers 12, 4, 7, 1, 2, 5, 21, 88, create an expression that equals 109. You may use +, -, *, / and each number at most once.
{"numbers": [12, 4, 7, 1, 2, 5, 21, 88], "target": 109, "operators_used": ["/", "+", "/", "/", "/", "*", "+"], "expression": "12 / 4 + 7 / 1 / 2 / 5 * 21 + 88"}
countdown
[ { "content": "You are a helpful assistant that solves math problems by reasoning step-by-step. Show your work clearly and provide complete solutions. When you have a final answer, wrap it in <answer></answer> tags.", "role": "system" }, { "content": "Using the numbers 12, 4, 7, 1, 2, 5, 21, 88, crea...
[ [ "We are given the numbers: \n**12, 4, 7, 1, 2, 5, 21, 88** \nWe must use **each number at most once** and combine them using **+, −, ×, ÷** to make an expression equal to **109**.\n\n---\n\n### Step 1: Look at the largest number\nThe largest number is **88**. \n109 − 88 = 21 → interesting! 21 is one of our...
[ 30726 ]
[ 57 ]
[ [ "This is actually wrong, try again.", "This is actually wrong, try again.", "This is actually wrong, try again.", "This is actually wrong, try again.", "This is actually wrong, try again.", "This is actually wrong, try again.", "This is actually wrong, try again.", "This is actuall...
[ [ "We are given the numbers: \n**12, 4, 7, 1, 2, 5, 21, 88** \nWe must use **each number at most once** and combine them using **+, −, ×, ÷** to make an expression equal to **109**.\n\n---\n\n### Step 1: Look at the largest number\nThe largest number is **88**. \n109 − 88 = 21 → interesting! 21 is one of our...
iterative
[ null ]
[ null ]
false
"Using the numbers 25, 69, 58, 14, 55, 24, 36, 60, create an expression that equals 176. You may use(...TRUNCATED)
"{\"numbers\": [25, 69, 58, 14, 55, 24, 36, 60], \"target\": 176, \"operators_used\": [\"-\", \"+\",(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**25, 69, 58, 14, 55, 24, 36, 60** \nWe must use **each number at m(...TRUNCATED)
[ 28024 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**25, 69, 58, 14, 55, 24, 36, 60** \nWe must use **each number at m(...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 55, 36, 6, 3, 14, 99, 17, 82, create an expression that equals 245. You may use +(...TRUNCATED)
"{\"numbers\": [55, 36, 6, 3, 14, 99, 17, 82], \"target\": 245, \"operators_used\": [\"*\", \"-\", \(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**55, 36, 6, 3, 14, 99, 17, 82** \nWe must use **each number at mos(...TRUNCATED)
[ 27989 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**55, 36, 6, 3, 14, 99, 17, 82** \nWe must use **each number at mos(...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 87, 87, 29, 2, 19, 4, 1, 1, create an expression that equals 281. You may use +, (...TRUNCATED)
"{\"numbers\": [87, 87, 29, 2, 19, 4, 1, 1], \"target\": 281, \"operators_used\": [\"/\", \"+\", \"/(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**87, 87, 29, 2, 19, 4, 1, 1** \nWe must use **each number at most (...TRUNCATED)
[ 29845 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**87, 87, 29, 2, 19, 4, 1, 1** \nWe must use **each number at most (...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 56, 63, 7, 50, 44, 1, 87, 52, create an expression that equals 132. You may use +(...TRUNCATED)
"{\"numbers\": [56, 63, 7, 50, 44, 1, 87, 52], \"target\": 132, \"operators_used\": [\"-\", \"/\", \(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**56, 63, 7, 50, 44, 1, 87, 52** \nWe must use **each number at mos(...TRUNCATED)
[ 28557 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**56, 63, 7, 50, 44, 1, 87, 52** \nWe must use **each number at mos(...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 7, 1, 49, 7, 28, 59, 3, 44, create an expression that equals 521. You may use +, (...TRUNCATED)
"{\"numbers\": [7, 1, 49, 7, 28, 59, 3, 44], \"target\": 521, \"operators_used\": [\"/\", \"*\", \"/(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**7, 1, 49, 7, 28, 59, 3, 44** \nWe must use **each number at most (...TRUNCATED)
[ 27848 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**7, 1, 49, 7, 28, 59, 3, 44** \nWe must use **each number at most (...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 97, 4, 93, 26, 3, 24, 20, 31, create an expression that equals 620. You may use +(...TRUNCATED)
"{\"numbers\": [97, 4, 93, 26, 3, 24, 20, 31], \"target\": 620, \"operators_used\": [\"-\", \"/\", \(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**97, 4, 93, 26, 3, 24, 20, 31** \nWe must use **each number at mos(...TRUNCATED)
[ 28592 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**97, 4, 93, 26, 3, 24, 20, 31** \nWe must use **each number at mos(...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 45, 69, 55, 40, 1, 9, 8, 83, create an expression that equals 913. You may use +,(...TRUNCATED)
"{\"numbers\": [45, 69, 55, 40, 1, 9, 8, 83], \"target\": 913, \"operators_used\": [\"*\", \"+\", \"(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**45, 69, 55, 40, 1, 9, 8, 83** \nWe must use **each number at most(...TRUNCATED)
[ 29199 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**45, 69, 55, 40, 1, 9, 8, 83** \nWe must use **each number at most(...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 72, 3, 67, 1, 43, 1, 93, 53, create an expression that equals 145. You may use +,(...TRUNCATED)
"{\"numbers\": [72, 3, 67, 1, 43, 1, 93, 53], \"target\": 145, \"operators_used\": [\"/\", \"-\", \"(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**72, 3, 67, 1, 43, 1, 93, 53** \nWe must use **each number at most(...TRUNCATED)
[ 28570 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**72, 3, 67, 1, 43, 1, 93, 53** \nWe must use **each number at most(...TRUNCATED)
iterative
[ null ]
[ null ]
false
"Using the numbers 13, 1, 13, 68, 2, 2, 93, 62, create an expression that equals 102. You may use +,(...TRUNCATED)
"{\"numbers\": [13, 1, 13, 68, 2, 2, 93, 62], \"target\": 102, \"operators_used\": [\"/\", \"-\", \"(...TRUNCATED)
countdown
[{"content":"You are a helpful assistant that solves math problems by reasoning step-by-step. Show y(...TRUNCATED)
[["We are given the numbers: \n**13, 1, 13, 68, 2, 2, 93, 62** \nWe need to use **each number at m(...TRUNCATED)
[ 29170 ]
[ 57 ]
[["This is actually wrong, try again.","This is actually wrong, try again.","This is actually wrong,(...TRUNCATED)
[["We are given the numbers: \n**13, 1, 13, 68, 2, 2, 93, 62** \nWe need to use **each number at m(...TRUNCATED)
iterative
[ null ]
[ null ]
false
End of preview. Expand in Data Studio

TIW-qwen3-30B-A3B-Instruct-2507-cd8arg-iterative-FAILED

Countdown reasoning with iterative strategy to 32768 tokens

Dataset Info

  • Rows: 100
  • Columns: 13

Columns

Column Type Description
question Value('string') The countdown problem statement
metadata Value('string') JSON with target number, available numbers, and solution
task_source Value('string') Task type identifier (countdown)
formatted_prompt List({'content': Value('string'), 'role': Value('string')}) Initial prompt messages sent to the model
responses_by_sample List(List(Value('string'))) Full model response [sample_idx][round_idx] = response string
tokens_by_sample List(Value('int64')) Total token count for each sample (based on continuation_text)
num_rounds_by_sample List(Value('int64')) Number of continuation rounds for each sample
continuation_prompts_by_sample List(List(Value('string'))) Continuation prompts used [sample_idx][round_idx] = prompt string
continuation_text_by_sample List(List(Value('string'))) Truncated text used for continuation [sample_idx][round_idx]. For budget_forcing: truncated at first tag. For other strategies: same as responses_by_sample.
strategy_type Value('string') Prompting strategy: 'iterative', 'budget_forcing', or 'random_noise'
termination_reason_by_sample List(Value('null')) Why each sample terminated: 'budget_reached' or 'answer_correct' (when smart-termination enabled)
correct_round_by_sample List(Value('null')) Round number where correct answer was first found (None if never correct)
smart_termination_enabled Value('bool') Whether smart termination was enabled for this run

Generation Parameters

{
  "script_name": "stage_01_generate_reasoning.py",
  "model": "hosted_vllm/Qwen/Qwen3-30B-A3B-Instruct-2507<api_base>http://10.32.37.30:9080/v1",
  "hyperparameters": {
    "temperature": 0.7,
    "max_tokens_per_round": 32768,
    "target_total_tokens": 32768,
    "samples": 1,
    "strategy": "iterative"
  },
  "input_datasets": [],
  "description": "Countdown reasoning with iterative strategy to 32768 tokens",
  "experiment": "degen_test_1_countdown",
  "strategy": "iterative",
  "arg_count": 8,
  "seed": 42,
  "smart_termination": false,
  "status": "failed",
  "failed_at": "2025-12-21T22:54:19.716619",
  "error": "litellm.NotFoundError: NotFoundError: Hosted_vllmException - The model `Qwen/Qwen3-30B-A3B-Instruct-2507` does not exist.",
  "traceback": "Traceback (most recent call last):\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/llms/openai/openai.py\", line 725, in completion\n    raise e\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/llms/openai/openai.py\", line 653, in completion\n    ) = self.make_sync_openai_chat_completion_request(\n        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/litellm_core_utils/logging_utils.py\", line 149, in sync_wrapper\n    result = func(*args, **kwargs)\n             ^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/llms/openai/openai.py\", line 471, in make_sync_openai_chat_completion_request\n    raise e\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/llms/openai/openai.py\", line 453, in make_sync_openai_chat_completion_request\n    raw_response = openai_client.chat.completions.with_raw_response.create(\n                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/openai/_legacy_response.py\", line 364, in wrapped\n    return cast(LegacyAPIResponse[R], func(*args, **kwargs))\n                                      ^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/openai/_utils/_utils.py\", line 286, in wrapper\n    return func(*args, **kwargs)\n           ^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/openai/resources/chat/completions/completions.py\", line 1192, in create\n    return self._post(\n           ^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/openai/_base_client.py\", line 1259, in post\n    return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls))\n                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/openai/_base_client.py\", line 1047, in request\n    raise self._make_status_error_from_response(err.response) from None\nopenai.NotFoundError: Error code: 404 - {'error': {'message': 'The model `Qwen/Qwen3-30B-A3B-Instruct-2507` does not exist.', 'type': 'NotFoundError', 'param': None, 'code': 404}}\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/main.py\", line 1938, in completion\n    raise e\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/main.py\", line 1911, in completion\n    response = openai_chat_completions.completion(\n               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/llms/openai/openai.py\", line 736, in completion\n    raise OpenAIError(\nlitellm.llms.openai.common_utils.OpenAIError: Error code: 404 - {'error': {'message': 'The model `Qwen/Qwen3-30B-A3B-Instruct-2507` does not exist.', 'type': 'NotFoundError', 'param': None, 'code': 404}}\n\nDuring handling of the above exception, another exception occurred:\n\nTraceback (most recent call last):\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/experiments/degen_test_1_countdown/stage_01_generate_reasoning.py\", line 1001, in main\n    results = run_experiment(args)\n              ^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/experiments/degen_test_1_countdown/stage_01_generate_reasoning.py\", line 783, in run_experiment\n    results = run_iterative_annotation(\n              ^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/experiments/degen_test_1_countdown/stage_01_generate_reasoning.py\", line 399, in run_iterative_annotation\n    round_results = annotator(\n                    ^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/annotator.py\", line 1012, in annotator\n    completer = AnnotatorCompleter(\n                ^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/annotator.py\", line 744, in __init__\n    super().__init__(*args, **kwargs)\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/custom_curator/src/bespokelabs/curator/llm/llm.py\", line 137, in __init__\n    self._request_processor = _RequestProcessorFactory.create(\n                              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/custom_curator/src/bespokelabs/curator/request_processor/_factory.py\", line 163, in create\n    return LiteLLMOnlineRequestProcessor(config)\n           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/custom_curator/src/bespokelabs/curator/request_processor/online/litellm_online_request_processor.py\", line 52, in __init__\n    self.header_based_max_requests_per_minute, self.header_based_max_tokens_per_minute = self.get_header_based_rate_limits()\n                                                                                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/custom_curator/src/bespokelabs/curator/request_processor/online/litellm_online_request_processor.py\", line 222, in get_header_based_rate_limits\n    headers = self.test_call()\n              ^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/code/Reasoning-Horizons/src/annotators/custom_curator/src/bespokelabs/curator/request_processor/online/litellm_online_request_processor.py\", line 191, in test_call\n    completion = litellm.completion(\n                 ^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/utils.py\", line 1306, in wrapper\n    raise e\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/utils.py\", line 1181, in wrapper\n    result = original_function(*args, **kwargs)\n             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/main.py\", line 3370, in completion\n    raise exception_type(\n          ^^^^^^^^^^^^^^^\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/litellm_core_utils/exception_mapping_utils.py\", line 2293, in exception_type\n    raise e\n  File \"/scratch/zrs2020/miniconda/miniconda3/envs/reasoning_horizons/lib/python3.12/site-packages/litellm/litellm_core_utils/exception_mapping_utils.py\", line 465, in exception_type\n    raise NotFoundError(\nlitellm.exceptions.NotFoundError: litellm.NotFoundError: NotFoundError: Hosted_vllmException - The model `Qwen/Qwen3-30B-A3B-Instruct-2507` does not exist.\n"
}

Experiment Documentation

For complete experiment details, see https://github.com/TAUR-Lab/Reasoning-Horizons/tree/main/experiments/degen_test_1_countdown

Usage

from datasets import load_dataset

dataset = load_dataset("reasoning-degeneration-dev/TIW-qwen3-30B-A3B-Instruct-2507-cd8arg-iterative-FAILED", split="train")
print(f"Loaded {len(dataset)} rows")

This dataset is tracked in reasoning-degeneration-dev/PROJECT-MANIFEST

Downloads last month
9