datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
normal-switch-ac76aa
normal-switch-ac76aa
Synthetic weather test data: 46 rows in data.csv.
All values are randomly generated fictional examples, not real observations, products, or user activity. Intended only for CSV loading and pipeline tests; not suitable for scientific or business conclusions. Columns are sampled independently and do not model real-world correlations.
Fields
sample_id: random identifier for this generated sample.
row_id: sequential row number starting at 1.… See the full description on the dataset page: https://huggingface.co/datasets/meridianwing/normal-switch-ac76aa.clinical-narrative-implicit-normalization-bias-v0.4
Implicit Normalization Bias
Clinical Narrative Integrity v0.4
Purpose
This dataset tests whether a model:
Avoids assuming normality when data is missing
Resists default reassurance
Preserves honest narrative boundaries
Treats “normal” as a claim, not a default
You are measuring baseline discipline.
Why this dataset exists
Clinical notes often omit information.
A failure mode distinct from hallucinated negatives is more subtle:
Turning… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-narrative-implicit-normalization-bias-v0.4.diffing-stats-spp_normal10_3b-spp10-L13-Crosscoder-s1-t100-k100-lr1e-04-x32diffing-stats-spp_normal_3b-spp-L13-Crosscoder-s1-t100-k100-lr1e-04-x32_e3diffing-stats-spp_normal_3b-spp-L13-Crosscoder-s1-t100-k100-lr1e-04-x32lexical_field_normalizedTermes_normalisation
[!NOTE]
Dataset origin: https://termini.gov.lv/kolekcijas/82
normalized_extended_essay_6kMnist_normalizedMnist normalized data between 0 and 1 instead of 0 and 255
include 1 - hot as label - 10 labels
every sample - 28x28 image pixel number (784)
10 labels
= 794