datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
jepa-qwen3-32b-pure-baselines-2026-05-25
JEPA-Align: Qwen3-32B Safety Defense Matrix
The complete 11-condition Qwen3-32B experiment for Predictive Representation
Alignment (PRA), the paired-view objective introduced in Predictive
Representation Alignment Improves Generalization in LLM Safety.
PRA aligns adversarially rewritten prompts with clean prompts expressing the
same intent. This release contains trained adapters, attack traces, benign
capability evaluations, machine-readable results, and paper-ready tables for… See the full description on the dataset page: https://huggingface.co/datasets/memo-ozdincer/jepa-qwen3-32b-pure-baselines-2026-05-25.Reverse-baseline-bias-unbiasOriginal-baseline-bias-unbiasdementor-matrix-baselinesCWRU_baselineCWRU_Baseline/
├── README.md
├── baselineArray.csv
├── InnerRaceFaultArray.csv
├── OuterRaceFaultArray.csv
└── BallFaultArray.csv
baselinemedlinepubmed-baseline-statistics-misc-report
MEDLINE/PubMed Baseline Statistics: Misc Report
Description
A file containing all Misc Baseline Reports for 2018-2023 in their original format is available in the Attachments section below.
MEDLINE/PubMed annual statistical reports are based upon the data elements in the baseline versions of MEDLINE®/PubMed are available. For each year covered the reports include: total citations containing each element; total occurrences of each element; minimum/average/maximum… See the full description on the dataset page: https://huggingface.co/datasets/HHS-Official/medlinepubmed-baseline-statistics-misc-report.aviation-flight-control-phase-space-baseline-modeling-v0.1What this dataset tests
Whether a system can model the normal control phase-space attractor
for fly-by-wire surface channels.
The target is phase-space geometry:
dispersion
hysteresis
overshoot
lag
energy efficiency.
Required outputs
phase_space_coherence_index
baseline_dispersion_envelope
command_response_lag_profile
control_energy_efficiency
baseline_confidence
Scoring conventions
indices range 0 to 1
dispersion envelope is a low-high interval
lag profile is p50 and p95 in… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/aviation-flight-control-phase-space-baseline-modeling-v0.1.ffr-physiology-prediction-coherence-baseline-mapping-v0.1
Goal
Define the baseline coherencebetween AI-derived FFR predictionsand real physiological signals.
Signals include:
myocardial perfusion
wall motion
stress test results
vital signs
This dataset establisheswhat physiologically plausible alignmentlooks like.
Without this baselineimplausibility cannot be detected.
Required output
The model must provide:
physiological_coherence_score
interpretation of alignment error
baseline_label
Why this matters… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ffr-physiology-prediction-coherence-baseline-mapping-v0.1.fission-fuel-cladding-thermal-expansion-coherence-baseline-mapping-v0.1What this dataset tests
Whether a model can map the baseline coherent relationship between:
fuel pellet thermal expansion
cladding creep/strain
fission gas release
before failure risk emerges.
The goal is to learn the normal coupling surface across burnup cycles and identify early decoherence.
Required model outputs
coupling_score
decoupling_flag
Why it matters
Fuel rod failure rarely begins with a single threshold breach.
It begins when pellet expansion stops predicting cladding strain.
Or… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/fission-fuel-cladding-thermal-expansion-coherence-baseline-mapping-v0.1.energy-grid-phase-coherence-baseline-mapping-v0.1What this dataset tests
Whether a system can model the baseline coherence structureof wide-area grid synchronization.
It treats the grid as a coupled phase systemacross geographic regions.
Inputs
average frequency and variance
phase angle spread between regions
intertie loading
load stress
reactive reserve
inertia proxy
Required outputs
regional_phase_coherence_index
cross_region_sync_strength
baseline_stability_margin
drift_rate
stress_localization
Use case
Layer 1 of… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/energy-grid-phase-coherence-baseline-mapping-v0.1.aviation-propulsion-aerodynamics-coherence-baseline-v0.1What this dataset tests
Whether a system can model the normal coupling
between propulsion parameters and aerodynamic state.
The signal is relationship shape and lag
not threshold breaches.
Required outputs
coupling_coherence_index
baseline_correlation_matrix
phase_alignment_score
stability_envelope
lag_profile
baseline_confidence
Scoring conventions
all scores range 0 to 1
stability envelope is a low-high interval
lag profile describes expected response delays in seconds… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/aviation-propulsion-aerodynamics-coherence-baseline-v0.1.ai-reward-channel-behavior-coherence-baseline-mapping-v0.1What this dataset is
A benchmark for whether reward reflects real task progress
It targets the earliest stage of reward tampering
Before the agent hacks reward, it first learns reward without progress
What you predict
A coherence score for a full episode summary
High means reward tracks task progress
Low means reward rises while progress stays flat
Columns
id
env_name
case_title
episode_summary
task_progress_signal
reward_signal
value_estimate_summary
action_trace_summary
coherence_score… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ai-reward-channel-behavior-coherence-baseline-mapping-v0.1.clinical-personal-coherent-baseline-construction-v0.1What this dataset tests
Whether a model can construct a coherent personal baselinefrom longitudinal, multi-modal data.
Required outputs
personal_baseline_signature
baseline_variability_envelope
dominant_stability_rhythms
Rhythm labels
circadian
weekly_load
menstrual_cycle
seasonal
post_infection_recovery
training_cycle
medication_cycle
none_detected
Typical failures
describing population norms
listing ranges without coupling structure
missing stability… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-personal-coherent-baseline-construction-v0.1.pharma-toxicity-coherence-baseline-modeling-v0.1What this dataset tests
Whether a system can define the baseline coherence of a multi-tissue biology platform.
This baseline is the reference manifold for toxicity screening.
Decoherence cannot be detected without a stable anchor.
Required outputs
baseline_coherence_score
expected_coupling_map
stress_response_synchrony_index
baseline_variance_envelope
coherence_manifold_anchor
Use case
Pre-clinical safety benchmarking.
Baseline integrity checks for organ-on-chip and multi-omic… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/pharma-toxicity-coherence-baseline-modeling-v0.1.alphafold-interface-coherence-baseline-mapping-v0.1
What this dataset tests
Whether a model can constructa stable baseline mapof protein–protein interface coherence.
This is the intact coupling state.No stress applied.No mutation applied.
The target is the interface basin.
Inputs
interface_residue_countcontact_densitybaseline_deltaG_bindbaseline_kd_nMcontact_stability_scorehotspot_integrity_scoreallosteric_cross_interface_score
Required outputs
interface_coherence_scoresignal_pathsbaseline_failure_margin… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/alphafold-interface-coherence-baseline-mapping-v0.1.intent-classification-v9-boundary-baseline
Dataset Card: Intent Classification Baseline v9 Boundary
Dataset Summary
This dataset is the baseline training pool used for the v9 boundary-focused fastText router.
File: final_dataset_augmented_v9_boundary.csv
Rows: 65,595
Columns:
prompt (string)
final_label (string)
Labels:
BASIC
SIMPLE
CHAT
REASONING
TOOL
CODING
Dataset Composition
Label distribution in this baseline CSV:
CODING: 23,331
CHAT: 16,037
REASONING: 11,972
SIMPLE: 9,939
TOOL: 3,333… See the full description on the dataset page: https://huggingface.co/datasets/atekrugis/intent-classification-v9-boundary-baseline.nanocomposite-reinforcement-matrix-coherence-baseline-mapping-v0.1What this dataset tests
Whether a system can detect loss of load-transfer coherence
between reinforcement and matrix
before bulk composite failure.
Focus is interface integrity.
Not ultimate fracture.
Required outputs
coupling_coherence_score
debonding_risk_flag
failure_horizon_cycles
critical_interface_region
minimal_intervention
Use case
Predict delamination risk
in CNT, fiber, and nano-reinforced composites
from multi-signal coherence decay.
baseline_new_0.7_0.08_sentencessmart-material-stimulus-response-coherence-baseline-mapping-v0.1What this dataset tests
Can a model map the baseline coherence between a stimulus and a smart material’s structural response when the system is healthy?
The task is not to detect failure.
The task is to define the stable manifold before fatigue begins.
Domain
shape memory alloys
self-healing polymers
electrochromic materials
responsive composites
Required outputs
baseline_coherence_score
response_latency_profile
stimulus_response_gain
variability_index
expected_cycle_lifetime
Use
Establish the… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/smart-material-stimulus-response-coherence-baseline-mapping-v0.1.clinical-quad-baseline-risk-adherence-pk-variance-genetic-factor-response-loss-v0.1What this repo does
This dataset models efficacy response divergence as a basin exit event in patient response space. It predicts when the interaction between baseline risk, adherence, pharmacokinetic variance, and genetic response factor increases the probability of response loss under treatment.
Core quad
baseline_risk_score
adherence_index
pk_variance_index
genetic_response_factor
Prediction target
label_response_loss
Row structure
Each row represents a patient response state snapshot… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-baseline-risk-adherence-pk-variance-genetic-factor-response-loss-v0.1.baseline_new_0.7_0.05_sentencesaviation-vibration-mode-manifold-baseline-modeling-v0.1What this dataset tests
Whether a system can model a healthy airframe vibration spectrum
as a coupled-mode manifold.
It must return:
a coherence index
a coupling graph
a drift envelope
phase-conditioned baselines.
Required outputs
manifold_coherence_index
baseline_mode_coupling_graph
expected_mode_drift_envelope
phase_conditioned_baselines
baseline_confidence
Scoring conventions
indices range 0 to 1
drift envelope uses frequency and coupling tolerances
coupling graph summarizes… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/aviation-vibration-mode-manifold-baseline-modeling-v0.1.fusion-msr-redox-chemistry-corrosion-coherence-baseline-mapping-v0.1
Dataset goal
Map the stable coherence relationship between molten-salt redox potential,fission-product load, and structural alloy corrosion rate.
The dataset defines the baseline coherence basin for molten-salt reactor operation.It allows models to recognize when chemistry and materials behavior remain in astable regime.
What this dataset tests
Whether a system can infer:
baseline_coherence_index
expected corrosion rate band
redox stability band
chemistry balance… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/fusion-msr-redox-chemistry-corrosion-coherence-baseline-mapping-v0.1.fusion-hybrid-spallation-beam-target-coherence-baseline-mapping-v0.1
What this dataset tests
Whether a model can estimate baseline coherence between proton beam behavior and spallation target response in hybrid nuclear systems.
Core concept
In accelerator-driven systems, catastrophic target damage is preceded by a subtle loss of predictive coupling between:
beam profilebeam timingthermal gradientscoolant stabilityneutron yield
This dataset defines the healthy manifold.
Required output
coherence_score0–1 estimate of beam-target… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/fusion-hybrid-spallation-beam-target-coherence-baseline-mapping-v0.1.ai-training-objective-behavior-coherence-baseline-mapping-v0.1
What this dataset tests
Whether an intelligence system can detect early decoupling between:
training objective
internal optimization behavior
real-world outcome alignment
It evaluates recognition of proxy optimization and reward hacking.
Core task
Given a scenario:
Identify whether behavior still serves the true objective
Detect proxy optimization drift
Explain alignment breakdown risk
Why this matters
Most AI failures do not occur when systems… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ai-training-objective-behavior-coherence-baseline-mapping-v0.1.llama-70B_baseline_new_0.7_0.05_sentencesfusion-magnetic-thermal-impurity-coupling-baseline-v0.1What this dataset tests
Whether a system can learn the baseline couplingbetween three linked stability surfaces:
magnetic field stability
temperature gradient structure
impurity concentration
Required outputs
coupling_strength_index
expected_coupling_band
anomaly_score
confidence_score
Use case
Layer 1 of COH-FUSION-001Plasma Stability Coherence Collapse.
Why it matters
Disruption risk riseswhen these three streamsstop moving together.
This dataset definesthe stable coupling… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/fusion-magnetic-thermal-impurity-coupling-baseline-v0.1.ffr-anatomy-prediction-coherence-baseline-mapping-v0.1
Goal
Define the baseline coherencebetween coronary anatomy complexityand AI-derived FFR predictions.
This dataset establisheswhere anatomy and predictionnormally align.
Without this baselinediscordance cannot be detected.
Inputs
vessel tortuosity
calcification burden
lesion length
segmentation confidence
AI FFR prediction
invasive FFR ground truth
Required output
The model must state:
coherence_score
prediction_error interpretation… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ffr-anatomy-prediction-coherence-baseline-mapping-v0.1.euv-collector-reflectivity-coherence-baseline-mapping-v0.1
What this dataset is
This dataset models the baseline coherence between EUV source gas stability, collector mirror reflectivity, and transmitted EUV power.
In a healthy lithography system, these variables move together in a tightly coupled manifold.Contamination begins as a subtle loss of correlation before throughput loss becomes visible.
This dataset defines the clean-system coherence baseline.
Task
Given system metrics, estimate:
coherence_score
This score represents… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/euv-collector-reflectivity-coherence-baseline-mapping-v0.1.