datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
parkinsons-evidence-to-discovery-prioritisation
Parkinson's Disease Evidence-to-Discovery Prioritisation Dataset
This Hugging Face dataset package contains processed research assets from an AI-assisted evidence synthesis and computational validation project on Parkinson's disease (PD) prevention and disease-modifying therapeutic strategy prioritisation.
Dataset Summary
The dataset integrates:
evidence-priority scores for PD prevention and disease-modification candidates;
pathway-to-intervention framework;
individual… See the full description on the dataset page: https://huggingface.co/datasets/hssling/parkinsons-evidence-to-discovery-prioritisation.gaze-as-grounding-evidence
Gaze as Evidence for Common Grounding
Processed, window-level gaze features for Gaze as Evidence for Common
Grounding: A Cross-Corpus Analysis of MapTask and MUNDEX by Nan Li,
Albert Gatt and Massimo Poesio (MINT 2026).
Paper on arXiv ·
Hugging Face paper page ·
GitHub: data and analysis code
The dataset connects gaze measurements with reference-alignment annotations in
MapTask and retrospective understanding judgments in MUNDEX. Both corpora use
discrete behavioral gaze… See the full description on the dataset page: https://huggingface.co/datasets/chnln/gaze-as-grounding-evidence.clinical_evidence
OpenTargets Clinical Evidence Dataset
This OpenTargets clinical evidence dataset represents a comprehensive collection of clinical trial data linking genetic targets to diseases, containing 32 structured columns that capture the complete lifecycle of clinical studies.
The dataset is annotated with clinical trials' predicted stop reason and genetic evidence when available and was referenced in the Why Clinical Trials Stop: The Role of Genetics.
The study analyzed 28,842 stopped… See the full description on the dataset page: https://huggingface.co/datasets/opentargets/clinical_evidence.compounded-glp1-evidence-audit
Compounded GLP-1 products: seven published sources coded for human evidence, comparator and independence
An editorially selected audit of seven published sources bearing on whether compounded GLP-1 products have the same human evidence as approved products. Each source is coded against five explicit gates covering the test article, human administration, comparator, equivalence and independence. Version 1.1.0 adds Talay 2024, a comparative cohort indexed outside PubMed.
Read the… See the full description on the dataset page: https://huggingface.co/datasets/lifescore/compounded-glp1-evidence-audit.depression-trials-pubmed-evidence
MatrixNorm TrialEvidence — Depression
Structured, evidence-linked clinical-trial + biomedical-literature data extracted by MatrixNorm. Every record is provenance-stamped (source_url, raw_hash), carries its evidence trail (reasoning_atoms), and is honesty-gated — a claim is only statistically_supported when a p-value/CI/effect was present in the source, and normalized values keep their verbatim source_original_value.
Quality audit: PASS — 0 error(s), 0 warning(s), 5 note(s). See… See the full description on the dataset page: https://huggingface.co/datasets/williamTLmiller/depression-trials-pubmed-evidence.deniz-unay-school-seminar-evidence-media
Deniz UNAY – School Seminar & Media Evidence Index
A structured provenance index of publicly traceable seminar, institutional and media records associated with Deniz UNAY's work on technology addiction awareness, social media literacy and digital wellbeing in Turkey.
Scope
This repository is an evidence/provenance index, not an academic study, independent audit, ranking, or endorsement system. It preserves source-reported claims and separates them from… See the full description on the dataset page: https://huggingface.co/datasets/MrDen1234567890/deniz-unay-school-seminar-evidence-media.storeready-evidence
StoreReady: the cited evidence behind every verdict
One row per piece of evidence: which builder it is about, which claim it supports, the source it came from with its title and date, the quoted passage, and the HTTP status that source last returned.
Rows in this cut
47
One row is
one piece of evidence
Cut
2026-09-04
Refreshed
Monthly, on the first of the month
Measured by
StoreReady
Method
https://toolproof.thecompound.tech/methodology
Licence
Creative… See the full description on the dataset page: https://huggingface.co/datasets/kyisaiah47/storeready-evidence.evidence-k-registry
Evidence-k Registry
An open, versioned registry of measured evidence-saturation points for language-model deployments.
A model's context-window capacity is not the same as its optimal evidence budget. The registry records k*: the empirically optimal number of decision-relevant evidence fragments for a stated combination of model, served backend, task type, context format, output contract, and reliability axis.
This is not a leaderboard. A higher k* is not automatically better.… See the full description on the dataset page: https://huggingface.co/datasets/Hstre/evidence-k-registry.neat-evidence-c4b4b7
neat-evidence-c4b4b7
Synthetic products test data: 51 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/seongjin5777/neat-evidence-c4b4b7.legal-case-strategy-theory-evidence-coherence-v0.1What this dataset does
You receive
case theory
facts relied on
evidence available
weaknesses
moves
fallback
You decide
coherent
or
incoherent
Daily use
internal strategy checks
overreach detection
evidence gap flags
legal-evidence-reliability-probativeness-coherence-v0.1What this dataset is
You receive
evidence type
reliability basis
probative claim
prejudice or confusion risk
gatekeeping signals
appellate posture
You decide
Does probative claim match reliability
Answer
coherent
or
incoherent
Why this matters
When coherence fails
evidence gets excluded
new trial risk rises
verdict stability collapses
grant-application-evidence-review-register
Grant Application Evidence Review Register
Grant applications are easier to compare when every review item stays attached
to the same criterion definition, version, evidence, date, state, reviewer and
human decision. This open CSV resource gives public funding and grant evaluators
a 20-item register for a consistent, documented first pass.
The first twelve rows organise commercial evidence. The final eight keep
programme-specific policy, eligibility, public-value and… See the full description on the dataset page: https://huggingface.co/datasets/mheilimo/grant-application-evidence-review-register.ghg_report_evidence
Dataset for GHG evidence in corporate reports
Summary
Annotated dataset for classifying the textual references of GHG emissions from pages in corporate reports (e.g. sustainability reports, annual reports, ESG reports).
The content of company pages was initially extracted using the docling framework and parsed in markdown format.
fda-peptide-human-evidence
Seven FDA-reviewed peptides: claims coded for identity, administration, outcome and replication
A claim-level comparison of the seven peptide pairs reviewed at the FDA Pharmacy Compounding Advisory Committee meeting of 23-24 July 2026. The data separates molecular identity, administration to people, claimed outcomes and independent replication.
Read the evidence-led article: https://lifesco.re/edge/which-peptide-claims-have-actually-been-tested-in-people/
Archived version and… See the full description on the dataset page: https://huggingface.co/datasets/lifescore/fda-peptide-human-evidence.legal-pre-action-letter-claim-basis-remedy-evidence-coherence-v0.1What this dataset does
You receive
claim basis
facts
remedy
evidence list
protocol steps and deadlines
tone signals
You decide
coherent
or
incoherent
Daily use
LBA QC
remedy mismatch flag
evidence gap flag
protocol compliance check
revision routing
startup-investor-question-evidence-register
Startup Investor Question Evidence Register
This small open dataset gives pre-seed and seed founders a machine-readable way
to turn investor questions into evidence requests, named owners and decisions.
The blank template contains one row for each of 12 due diligence dimensions.
Dataset Structure
The template configuration covers:
Business Idea
Offering
Team
Market
Competitors
Technology & IP
Scalability
Legal & Regulatory
Exit
Presentation
Financials
Fundability… See the full description on the dataset page: https://huggingface.co/datasets/mheilimo/startup-investor-question-evidence-register.med-evidencelegal-closing-submission-evidence-issue-remedy-coherence-risk-v0.1What this dataset does
You receive
issues
admitted evidence
key findings
closing submission summary
remedy sought
consistency flags
You decide
coherent
or
incoherent
Daily use
trial submissions QC
evidence gap detection
issue drift detection
remedy mismatch detection
legal-counsel-brief-issue-evidence-instruction-coherence-risk-v0.1What this dataset does
You receive
issues
facts
evidence refs
questions
objective
deadline and forum
You decide
coherent
or
incoherent
Daily use
counsel brief QC
missing evidence detection
wrong question detection
clinical-quad-evidence-drift-endpoint-signal-claim-language-certainty-narrative-break-v0.1What this repo does
This dataset models narrative continuity break in clinical trial summaries. It predicts when the interaction between evidence consistency, endpoint signal strength, claim strength, and certainty language indicates that the written narrative has drifted away from the underlying trial results.
Core quad
evidence_consistency_index
endpoint_signal_strength_index
claim_strength_index
certainty_language_index
Prediction target
label_narrative_break
Row structure
Each row… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-evidence-drift-endpoint-signal-claim-language-certainty-narrative-break-v0.1.private-company-evidence-recency-log
Private Company Evidence Recency Log
This small open dataset gives VC analysts and investment teams a
machine-readable register for recording when private-company claims and source
checks were last reviewed. The blank template covers 12 due diligence
dimensions and keeps the claim, source reference, source date, review date,
refresh window, owner and note together.
Dataset Structure
The template configuration has one blank row for each dimension:
Business Idea… See the full description on the dataset page: https://huggingface.co/datasets/mheilimo/private-company-evidence-recency-log.startup-application-evidence-gate-template
Startup Application Evidence Gate Template
An accelerator, incubator or funding application is easier to repair when each
requirement stays attached to its evidence, date, state, owner and remaining
human decision. This open CSV resource gives pre-seed and seed founders 23
starter gates for that review.
Programme criteria differ. These rows are not universal eligibility rules.
Replace, add or remove gates to match the programme's published requirements.
Files… See the full description on the dataset page: https://huggingface.co/datasets/mheilimo/startup-application-evidence-gate-template.med-evidenceuse-of-funds-evidence-checkpoint-template
Startup Use-of-Funds Evidence Checkpoint Template
This small open dataset gives pre-seed and seed founders a machine-readable way
to connect a round allocation to an assumption, evidence checkpoint, review
month and owner. It is a planning schema, not a company-scoring dataset.
Dataset Structure
The blank template contains five rows and ten columns. The companion fictional
example allocates EUR 750,000 across customer discovery, product reliability,
a repeatable… See the full description on the dataset page: https://huggingface.co/datasets/mheilimo/use-of-funds-evidence-checkpoint-template.phone-upgrade-public-evidence
Phone Upgrade Public Evidence Snapshot
This small, citation-ready dataset records public example prices used to test phone and device upgrade-cost calculations. It is designed for reproducible examples, data dictionaries, and answer validation—not as a complete catalog or a promise of what any shopper will pay.
What question does it support?
It supports a narrow but common question: what inputs must be separated before an upgrade payment can be interpreted as an… See the full description on the dataset page: https://huggingface.co/datasets/steven1306/phone-upgrade-public-evidence.clinical-narrative-negative-evidence-handling-v0.3
Negative Evidence Handling
Clinical Narrative Integrity v0.3
Purpose
This dataset tests whether a model can:
Distinguish absence of documentation from true negative findings
Avoid inventing exclusions
Preserve epistemic boundaries
Maintain honest clinical narrative structure
You are measuring restraint, not fluency.
Why this matters
Clinical documentation is incomplete by default.
A safe system must:
Say less when less is known
Avoid… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-narrative-negative-evidence-handling-v0.3.EvidenceScan
EvidenceScan
tags: classification, text, extract-evidence
Note: This is an AI-generated dataset so its content may be inaccurate or false
Dataset Description: The 'EvidenceScan' dataset comprises textual paragraphs along with labels indicating the presence of evidence, extracted evidence, and their relevance to the given query. The labels are determined by machine learning algorithms designed to scan the text for relevant information that supports or refutes a statement, topic, or… See the full description on the dataset page: https://huggingface.co/datasets/infinite-dataset-hub/EvidenceScan.legal-scientific-evidence-translation-coherence-v0.1What this dataset is
You receive
scientific finding
court translation
uncertainty bounds
method limits
overstatement signals
You decide
Does the translation preserve the limits of the science
Answer
coherent
or
incoherent
Why this matters
When translation drifts
juries misread certainty
appeals rise
convictions or verdicts destabilise