MME-SCI: A Comprehensive and Challenging Science Benchmark for Multimodal Large Language Models
Paper β’ 2508.13938 β’ Published β’ 1
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: CastError
Message: Couldn't cast
id: string
image: string
question: string
options: list<element: string>
child 0, element: string
answer: string
knowledge-source: string
course: string
-- schema metadata --
pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 881
to
{'id': Value('string'), 'image_1': Value('string'), 'image_2': Value('string'), 'image_3': Value('string'), 'image_4': Value('string'), 'image_5': Value('string'), 'question': Value('string'), 'options': List(Value('string')), 'answer': Value('string'), 'knowledge-source': Value('string'), 'course': Value('string')}
because column names don't match
Traceback: Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1404, in compute_config_parquet_and_info_response
fill_builder_info(builder, hf_endpoint=hf_endpoint, hf_token=hf_token, validate=validate)
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 596, in fill_builder_info
num_examples_and_sizes: list[tuple[int, int]] = thread_map(
^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/tqdm/contrib/concurrent.py", line 69, in thread_map
return _executor_map(ThreadPoolExecutor, fn, *iterables, **tqdm_kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/tqdm/contrib/concurrent.py", line 51, in _executor_map
return list(tqdm_class(ex.map(fn, *iterables, chunksize=chunksize), **kwargs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/tqdm/std.py", line 1169, in __iter__
for obj in iterable:
^^^^^^^^
File "/usr/local/lib/python3.12/concurrent/futures/_base.py", line 619, in result_iterator
yield _result_or_cancel(fs.pop())
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/concurrent/futures/_base.py", line 317, in _result_or_cancel
return fut.result(timeout)
^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/concurrent/futures/_base.py", line 456, in result
return self.__get_result()
^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/concurrent/futures/_base.py", line 401, in __get_result
raise self._exception
File "/usr/local/lib/python3.12/concurrent/futures/thread.py", line 59, in run
result = self.fn(*self.args, **self.kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 468, in retry_validate_get_num_examples_and_size
validate(pf)
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 534, in validate
raise TooBigRowGroupsError(
worker.job_runners.config.parquet_and_info.TooBigRowGroupsError: Parquet file has too big row groups. First row group has 483086943 which exceeds the limit of 300000000
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1815, in _prepare_split_single
for _, table in generator:
^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 691, in wrapped
for item in generator(*args, **kwargs):
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/parquet/parquet.py", line 106, in _generate_tables
yield f"{file_idx}_{batch_idx}", self._cast_table(pa_table)
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/parquet/parquet.py", line 73, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
id: string
image: string
question: string
options: list<element: string>
child 0, element: string
answer: string
knowledge-source: string
course: string
-- schema metadata --
pandas: '{"index_columns": [], "column_indexes": [], "columns": [{"name":' + 881
to
{'id': Value('string'), 'image_1': Value('string'), 'image_2': Value('string'), 'image_3': Value('string'), 'image_4': Value('string'), 'image_5': Value('string'), 'question': Value('string'), 'options': List(Value('string')), 'answer': Value('string'), 'knowledge-source': Value('string'), 'course': Value('string')}
because column names don't match
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1427, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 993, in stream_convert_to_parquet
builder._prepare_split(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1702, in _prepare_split
for job_id, done, content in self._prepare_split_single(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1858, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
id string | image_1 string | image_2 string | image_3 string | image_4 string | image_5 string | question string | options list | answer string | knowledge-source string | course string |
|---|---|---|---|---|---|---|---|---|---|---|
52011046_4 | null | null | null | null | null | Set\( M = \{ f(x), f'(x), f''(x), \cdots \} \), then the following can be expressions of\( f(x) \) are () | [
"\\(\\sin x\\)",
"\\(e^x\\)",
"\\(\\ln x\\)",
"\\(x^2 + 2x + 3\\)"
] | C | ιεδΈεΈΈη¨ι»θΎη¨θ― | math |
52011046_17 | null | null | null | null | null | Given the sets \( A = \left\{ x \mid x = \frac{k}{2} + \frac{1}{4}, k \in \mathbb{Z} \right\} \), \( B = \left\{ y \mid y = \frac{l}{4} + \frac{1}{2}, l \in \mathbb{Z} \right\} \), then ( ) | [
"\\( A \\cap B = \\emptyset \\)",
"\\( A = B \\)",
"\\( A \\not\\subseteq B \\)",
"\\( B \\not\\subseteq A \\)"
] | C | ιεδΈεΈΈη¨ι»θΎη¨θ― | math |
52011046_45 | iVBORw0KGgoAAAANSUhEUgAAAXcAAAFXCAYAAABZbA7IAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAAEnQAABJ0Ad5mH3gAAGAKSURBVHhe7Z0HmFRVtrb194a54+gY0FGMIAaCCgZUlCAgSVCSoICAJEEFARUJioKogCiSlKAIiAgIkiQIChIEEyKiGEDEgNk7zui9c+em/d93Oatmd1EN3U3VqRPW9zz7aaiuqj5n77W/s/aKBzmDwWAwxA5G7gaDwRBDGLkbDAZDDGHkbjAYDDGEkbvBYDDEEEbuBoPBEEMY... | null | null | null | null | <image_1>If the Venn graph representing the relationship between sets\( M \) and\( N \) is as shown in the graph, then\( M, \, N \) may be () | [
"\\( M = \\{0, 2, 4, 6\\}, \\, N = \\{4\\} \\)",
"\\( M = \\{x \\mid x^2 < 1\\}, \\, N = \\{x \\mid x > - 1\\} \\)",
"\\( M = \\{x \\mid y = \\log x\\}, \\, N = \\left\\{y \\mid y = e^x + \\frac{1}{e^x}\\right\\} \\)",
"\\( M = \\{(x, y) \\mid x^2 = y^2\\}, \\, N = \\{(x, y) \\mid y = x\\} \\)"
] | BCD | ιεδΈεΈΈη¨ι»θΎη¨θ― | math |
52011047-1_15 | null | null | null | null | null | Assuming that "Physics is good, mathematics is good" is a true proposition, then the following propositions are correct () | [
"Good physics may not necessarily be good mathematics",
"Good mathematics may not necessarily be good physics",
"Poor mathematics must be bad physics",
"Physics is bad, mathematics is bad"
] | BC | ιεδΈεΈΈη¨ι»θΎη¨θ― | math |
50948433-1_16 | iVBORw0KGgoAAAANSUhEUgAAAWIAAAFkCAYAAAAaBTFnAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAAEnQAABJ0Ad5mH3gAAETpSURBVHhe7Z0HuBTFtrbPNSdMqBhQBEVUUBG8BsziFSUoAornehRRMKJiDpgJKhzFHDCigooeVDyKAcVAMIFiRBSJegWPmPO5t/7nXf/U2Ay9957ae0L37O99nnpgd1f3THet/qZ61apVf3FCCCHKioRYCCHKjIRYCCHKjIRYCCHKjIRYCCHKjIRYCCHKjIRYCCHKjIRYCCHK... | iVBORw0KGgoAAAANSUhEUgAAAU8AAAF/CAYAAADJt0tfAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAAEnQAABJ0Ad5mH3gAADlaSURBVHhe7Z0J3FVz/sfHkC1knaxRlLRhNFSK7KNSSSgZS8oSjSRlNEqWlJRMoaJlEJGtlEjWGq0mWpTRZgnJUpR95vf/v3/d33Ge2733uc+59z733HM/79fr96rnbM99zvmd9/2t39/vjBBCiDIjeQohRAAkTyGECIDkKYQQAZA8hRAiAJKnEEIEQPIUQogASJ5CCBEAyVMI... | iVBORw0KGgoAAAANSUhEUgAAAJoAAACwCAYAAAD3yHdHAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAAEnQAABJ0Ad5mH3gAABM3SURBVHhe7Z0JtFXTH8eb1Ku0ECmFpHrpicxUKFSUNBhWJWSMUkloQin0TA2EWg2SWsjUYKoIaUCGhBIpQqJQIrP2///5uee5ve57vfveufudve/vs9ZZ9c65wxm+d+/9G/ZvlzCKYgEVmmIFFZpiBRWaYgUVmmIFFZpiBRWaYgUVmmIFFZpiBRWaYgUVmmIFFZpiBRWaYoV8... | iVBORw0KGgoAAAANSUhEUgAAAYMAAAGVCAYAAAAc652UAAAAAXNSR0IArs4c6QAAAARnQU1BAACxjwv8YQUAAAAJcEhZcwAAEnQAABJ0Ad5mH3gAAFOcSURBVHhe7Z0JvFXj/v9/l26IJHSVKfkhIpkyJZHMRZlK13gRFeGWoZQxyS1zmRLqphtXpJLI0IBKKkOh6C9TVCQkXe7vPv/7/q7zrNbZ9j6dszrD3mt93q/X8zq199rrnP2s53k+z/Ad/scJIYRIPRIDIYQQEgMhhBASAyGEEP9FYiCEEEJiIIQQQmIghBDiv0gMhBBCSAyE... | null | Given\( a > 0 \) and\( a \neq e \), then the image of the function\( f(x) = e^x - a \ln x \) may be () | [
"<image_1>",
"<image_2>",
"<image_3>",
"<image_4>"
] | BCD | ε½ζ°δΈε―Όζ° | math |
51626976-1_23 | null | null | null | null | null | Let\( m \) be a real number other than 0, and the function\( f(x) = \begin{cases}
|3x - 1|, & x \leq 2 \\
x^2 - 10x + 24, & x > 2
\end{cases} \), if the function\( F(x) = 2[f(x)]^2 - m f(x) \) has 7 zeros, then the value range of\( m \) is () | [
"\\((-2,0) \\cup (0,16)\\)",
"\\((0,16)\\)",
"\\((0,2)\\)",
"\\((-2,0) \\cup (0, +\\infty)\\)"
] | C | ε½ζ°δΈε―Όζ° | math |
51484881-1_23 | null | null | null | null | null | If the function\( y = f(x) \) satisfies a set\( A \) within the domain, and for any\( x \in A \),\( e^x [f(x) - e^x] \) is a constant\( a \), then\( f(x) \) is said to have the property\( M \) on\( A \). Let\( y = g(x) \) be a function with the property of\( M \) on the interval\([-2, 2]\), and for any\( x_1, x_2 \in [... | null | \([-e^4, e^4]\) | ε½ζ°δΈε―Όζ° | math |
52105556-1_25 | null | null | null | null | null | In\(\triangle ABC\), the edges bounded by the angles\(A, B, C\) are\(a, b, c\) respectively, and satisfy\[\frac{\sqrt{3}b \sin A}{1 + \cos B} = a\]. Find the value range of\(\sin A \sin C +\sin B \sin C + \sin B \sin A\). | null | \[\left[ \frac{3}{4}, \frac{9}{4} \right] | δΈθ§ε½ζ°δΈθ§£δΈθ§ε½’ | math |
52105556-1_18 | iVBORw0KGgoAAAANSUhEUgAAAVwAAAD1CAIAAACTENrpAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAgAElEQVR4nO29eUCM6///3z3TNlNN+77SqiRFKrRbUyhUOnZZsuREenOsWRLHkkjkZAmpROFQaFHanfaiRKV935tptvv+/XF/f30IaZmapevxF7Nc12umuZ/367qu1wIhCMIFAAAA/z8YZhsAAABYCyAKAADgO4AoAACA7wCiAAAAvgOIAgAA+A4gCmwPOD8CMBZuZhsAGAiCIEQisb6+vqampr6+3traWlJSEoIgBEHq6+tjYmIM... | null | null | null | null | <image_1>Given the function\( f(x) = A \sin(\omega x + \varphi) (A > 0, \omega > 0, 0 < \varphi < \pi) \), if the partial image of\( f(x) \) and its derivative function\(f'(x) \) is as shown in the figure, then () | [
"\\( f(x + \\pi) = f(x) \\)",
"Function\\( f(x) \\) monotonically decreases on\\(\\left( \\frac{\\pi}{12}, \\frac{7\\pi}{12} \\right)\\)",
"The image of\\(f'(x) \\) is symmetric about the center of the point\\(\\left( -\\frac{\\pi}{6}, 0 \\right)\\)",
"The maximum value of\\( f(x) + f'(x) \\) is\\(\\frac{5}{2... | AB | δΈθ§ε½ζ°δΈθ§£δΈθ§ε½’ | math |
52105556-1_9 | iVBORw0KGgoAAAANSUhEUgAAASkAAACSCAIAAACWgavXAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAf6ElEQVR4nO3dfTxU6f8/8Dkzg3F/N0wlCSMqRJtUbAkV+0FYpa3Ntps2EpUUm0ht5SbtJ1ZWN1t2P25S2BRlLe2ute26aRUpCU1EBuN+7mfO9f3jfH5+PdTupxvmjHE9/9rHOO15D/Oa65zrus51IQAAAgRBEkfEuwAImqJg9iAIHzB7EIQPmD0IwgfMHgThg4zXiVEUBQAgCEIkEgkEAgAAAICiKIlEQhAEO2BkZARFUQ0NDbyK... | iVBORw0KGgoAAAANSUhEUgAAASgAAACfCAIAAADF3dM3AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAVEklEQVR4nO3df1AU9f8H8Hvv3R533MFxyGEC8kN+Gr8MJEgEw9TKEPFH41RMYzWWZVTYNGWNozWjzQeVJvuF/dTSavJ3pqZFwEQToKAGchAIJxwXiHDAsnfc7e3u94/9fm4YxT6lt/c+4fX4z9313i/Xe+57971770U8z0sAAO5F4C4AgMkIggcABhA8ADCA4AGAAQQPAAwgeABgAMEDAAMIHgAYQPAAwACCBwAGogeP/69r/jh2... | iVBORw0KGgoAAAANSUhEUgAAARQAAACZCAIAAABhS1KcAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAgAElEQVR4nO29d3RbZZo/rqvebHXLtuQiuXcnthPSQwqBhBDKQjj0MGdgYGF3FmbZKbsDZ5ZZZnbZAzMLZ5aQ5cDMQMKBAAkhhOAU0h0cO3GJLblKcpGt3vu9vz+ec/Tz10W+uvdKcQZ9/kqurnQfS+/nffrzIhiG0TLIIIPkQb/RAmSQwc2KdJMHw7BgMHj9+vWMxsvgZgdl5EFR1O/322w2g8EwOTkZv+5wOPR6vdPpjLPlwIED... | iVBORw0KGgoAAAANSUhEUgAAARcAAACZCAIAAACKfOmfAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAgAElEQVR4nO29eVwT1/7/z2QhhCVsCatE9n1XAa3KIq4givvWKtb1qrUWaz/9WK92sbe1t9W21qXVarH9VKVaUcCKBRFlR1lkh7AKkoWQANkzc75/zO/mxwPQiySZBJ3nf0xmzjkznNdZ3ud93gcCABjg4OCoAUHXBcDBmfDgKsLBUReMVIQgiFgslslk2GSHg4MlGKmot7c3Nze3paUFm+xwcLBELRWB/6BQKBAEUV1UKBRSqVRl... | null | Part of the image of the function\( f(x) = \omega \sin\left(\omega^2x- \frac{\pi}{6} \right) + 1 \) may be () | [
"<image_1>",
"<image_2>",
"<image_3>",
"<image_4>"
] | ABC | δΈθ§ε½ζ°δΈθ§£δΈθ§ε½’ | math |
End of preview.
MME-SCI: A Comprehensive and Challenging Science Benchmark for Multimodal Large Language Models
MME-SCI is a comprehensive multimodal benchmark designed to evaluate the scientific reasoning capabilities of Multimodal Large Language Models (MLLMs). It addresses key limitations of existing benchmarks by focusing on multilingual adaptability, comprehensive modality coverage, and fine-grained knowledge point annotation.
π Key Features
- Multilingual Support: Covers 5 languages (Chinese, English, French, Spanish, Japanese) to assess cross-lingual scientific reasoning.
- Full Modality Coverage: Supports 3 evaluation modes (text-only, image-only, image-text hybrid) to test multimodal robustness.
- Multidisciplinary Focus: Includes 4 core subjects (mathematics, physics, chemistry, biology) at the high school level.
- Fine-Grained Knowledge Annotation: Annotates 63 specific knowledge points (e.g., "Magnetic Field" in physics, "Trigonometric Functions" in mathematics) for targeted performance analysis.
- High-Quality Data: 1,019 manually curated question-answer pairs, with 83.3% sourced from 2025 and 16.2% from 2024 to ensure novelty and avoid training data contamination.
π Dataset Details
| Attribute | Details |
|---|---|
| Total Samples | 1,019 question-answer pairs |
| Subjects & Distribution | - Mathematics: 284 - Physics: 368 - Chemistry: 151 - Biology: 216 |
| Languages | Chinese, English, French, Spanish, Japanese |
| Modality Modes | - Text-only: Pure textual questions - Image-only: Screenshot-based questions - Image-text hybrid: Combined visual and textual inputs |
| Question Types | Single-choice, multiple-choice, fill-in-the-blank (all with verifiable answers for automated scoring) |
| Knowledge Points | 63 fine-grained concepts (e.g., "Regulation of Plant Life Activities" in biology, "Basic Laws of Heredity" in genetics) |
π Evaluation Results
MME-SCI has been tested on 20 MLLMs (16 open-source, 4 closed-source), revealing significant challenges for existing models:
π Data Curation
The dataset was built through a rigorous 4-step process:
- Sample Filtering: High-difficulty questions selected by experts (top 0.1% in Gaokao).
- Data Digitization: Conversion to JSON format with OCR for image-based content.
- Language Transformation: Professional translation to 5 languages.
- Post-Audit: Cross-validation by 3 reviewers to ensure quality.
π Citation
If you use MME-SCI in your research, please cite:
@article{ruan2025mme,
title={MME-SCI: A Comprehensive and Challenging Science Benchmark for Multimodal Large Language Models},
author={Ruan, Jiacheng and Jiang, Dan and Gao, Xian and Liu, Ting and Fu, Yuzhuo and Kang, Yangyang},
journal={arXiv preprint arXiv:2508.13938},
year={2025}
}
π Resources
π§ Contact
For questions, reach out to jackchenruan@sjtu.edu.cn.
- Downloads last month
- 236