The dataset viewer is not available for this subset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
for split_generator in builder._split_generators(
~~~~~~~~~~~~~~~~~~~~~~~~~^
StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/harbor/harbor.py", line 171, in _split_generators
raise DataFilesNotFoundError("No task.toml or instruction.md files found")
datasets.exceptions.DataFilesNotFoundError: No task.toml or instruction.md files found
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/split_names.py", line 68, in compute_split_names_from_streaming_response
for split in get_dataset_split_names(
~~~~~~~~~~~~~~~~~~~~~~~^
path=dataset,
^^^^^^^^^^^^^
config_name=config,
^^^^^^^^^^^^^^^^^^^
token=hf_token,
^^^^^^^^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
info = get_dataset_config_info(
path,
...<6 lines>...
**config_kwargs,
)
File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Cross-study single-cell integration — task inputs
Temporary public hosting for the input bundle used in Terminal-Bench Science PR #2107, proposed in Discussion #1822.
The packaged cohort contains 30,262 human lung epithelial cells and 55,329 genes. It provides an uncorrected sparse UMI count matrix with permitted gene, study, donor and sample metadata. This repository contains no grading reference labels, saved embeddings, trained models, oracle code or grader.
| Path | Description |
|---|---|
cross-study-integration/input/cohort.h5ad |
Frozen solver input, 149,372,383 bytes |
cross-study-integration/input/dictionary.md |
Input fields and count conventions |
cross-study-integration/DATA_ATTRIBUTION.md |
Source, derivative scope and attribution |
cross-study-integration/release_manifest.json |
Preserved preparation manifest and payload SHA-256 checksums |
The cohort SHA-256 is fe78b0c617d369df558c2b9229e7aa22c58b2b4a6e98435383a2534b6765106b.
Source and attribution
Derived by Fei Wu, Buck Institute for Research on Aging, from the Human Lung Cell Atlas (Sikkema et al., Nature Medicine, 2023; doi:10.1038/s41591-023-02327-2), distributed through CELLxGENE. The subset, transformations and CC BY 4.0 source-policy basis are documented in DATA_ATTRIBUTION.md.
The attribution and release manifest are retained byte-for-byte from the September 30 preparation bundle. Their local-only status language describes that preparation snapshot. Author-approved public hosting occurred on October 6, 2026; it does not change the scientific input or its checksum.
Reproducible use and permanent hosting
Download from a full commit SHA, not a moving branch, and verify file checksums. Only cross-study-integration/input/* belongs in the solver image. This upload does not change the task's scientific criteria or establish a new build or execution result.
This personal dataset is temporary hosting during PR review. Before merge, maintainers will copy the bundle to harborframework/terminal-bench-science-lfs; the task must then use the organization repository's actual commit SHA and be revalidated. After that transfer is complete and the task is repointed, this development copy should be deleted or made private, following the contribution guidance.
- Downloads last month
- 40