datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
BiomniBench-DA
BiomniBench-DA
BiomniBench-DA is the data-analysis instantiation of BiomniBench, a process-level evaluation framework for LLM agents on real-world biomedical research tasks. Each task is a multi-step data analysis derived from a high-impact biomedical publication; agents are graded on the full analytical trajectory against an expert-authored rubric, not only the final answer.
This repository releases 50 of the 100 BiomniBench-DA tasks; the remaining 50 are held out as a private… See the full description on the dataset page: https://huggingface.co/datasets/phylobio/BiomniBench-DA.Eval1
Biomni-Eval1
To use this dataset, simply pip install biomni and then do:
from biomni.eval import BiomniEval1
# Initialize
evaluator = BiomniEval1()
# Get instance details
instance = evaluator.get_instance('gwas_causal_gene_opentargets', 0)
print(f"Prompt: {instance['prompt']}")
print(f"Answer: {instance['answer']}")
# Evaluate user answer
score = evaluator.evaluate('gwas_causal_gene_opentargets', 0, 'BRCA1')
print(f"Score: {score}") # 0.0 or 1.0
Important Notes… See the full description on the dataset page: https://huggingface.co/datasets/biomni/Eval1.biomnibench-da
OmicOS-BiomniBench Task Packages (Flat / Standardized)
Note: This directory is an improved, structural-flattened and
format-standardized derivative of the upstream
OmicOS-BiomniBench
benchmark. The original release stores each task as a directory of
heterogeneous files (instruction.md, task.toml, environment/,
tests/rubric.txt, ...). This package reformats every task into the
single-file, schema-versioned task-package layout consumed by our
downstream agent harness, so that all… See the full description on the dataset page: https://huggingface.co/datasets/wjixiang/biomnibench-da.biomnibench-organized
BioMniBench DA — Reorganized
A clean, manifest-driven reorganization of the
BioMniBench DA (Data Analysis) task suite,
shaped for use with the
biomnibench-adapter evaluation
harness and the native skill-learning loop that ships with it.
This Hugging Face repository hosts the metadata, evaluation rubric and data manifest for
all 50 tasks. The raw input data files (which total ~77 GB and originate upstream from
GEO/TCGA/cBioPortal/etc.) are not redistributed here — see
Getting… See the full description on the dataset page: https://huggingface.co/datasets/starpacker52/biomnibench-organized.biomni-data-lakecapsules-biomnibenchbiomni-datalakeBiomniBench-DA-sample
BiomniBench-DA-sample
A small representative sample of BiomniBench-DA, intended for reviewer inspection of dataset quality, structure, and per-task contents.
Full dataset: phylobio/BiomniBench-DA (50 released tasks; 50 additional held-out as a private contamination-resistant evaluation set).
How this sample was created
We selected 3 tasks from the 50-task release, one from each of three different disease areas, prioritizing small footprint so the sample can be… See the full description on the dataset page: https://huggingface.co/datasets/phylobio/BiomniBench-DA-sample.biomni-eval1
Biomni-Eval1 data
This directory is the local input for the native Biomni-Eval1 harness in
mega_eval/benchmarks/biomni_eval1.py.
Default mirror: ssswwwxxx/biomni-eval1
Official source: biomni/Eval1
Split: official val (433 instances across 10 tasks)
License: Apache-2.0
Local file: validation.parquet (ignored by Git)
Reproducibility: manifest.json pins the resolved Hub commit and file hash
Official evaluator reference: Biomni commit 400c1f366b96a35ca253e13c9b06c5076af41d65… See the full description on the dataset page: https://huggingface.co/datasets/ssswwwxxx/biomni-eval1.BiomniBench-AI4S
BiomniBench-AI4S — Agent Trajectories
Per-cell outputs from a horizontal comparison of 7 AI-for-Science agents on
the same 50 BiomniBench-DA biomedical
data-analysis tasks, under identical conditions — same model (deepseek-v4-pro),
same DeepSeek v4-pro rubric judge.
Leaderboard, harness, adapters, and analysis:
👉 https://github.com/omicverse/BiomniBench-AI4S
Layout
<backend>/<task>/
trace.md # the agent's structured analytical trace
answer.txt… See the full description on the dataset page: https://huggingface.co/datasets/omicverse/BiomniBench-AI4S.Biomni_Comparative_Experiments