datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
africa-synth-neonatal-sepsis-dataset-all
African Neonatal Sepsis Synthetic Dataset | Africa (Electric Sheep Africa metadata inventory)
Size category: 10K<n<100K - Formats: csv - Sector: health - Engineered by Electric Sheep Africa
TL;DR
This dataset is part of the Electric Sheep Africa catalog on Hugging Face. It is indexed for African data discovery with standardized metadata, loading guidance, provenance notes, and analyst-oriented context.
What This Dataset Covers
Health datasets… See the full description on the dataset page: https://huggingface.co/datasets/electricsheepafrica/africa-synth-neonatal-sepsis-dataset-all.SepsisPrediction_DatasetSepsisPrediction_Dataset
Dataforce Team - DATATHON Competition by RISTEK FASILKOM Universitas INDONESIA (2025)
clinical-sepsis-trajectory-instability-v0.1
clinical-sepsis-trajectory-instability-v0.1
What this dataset does
This dataset tests whether a model can classify sepsis trajectory instability from short clinical proxy sequences.
Each row describes a patient-like scenario across three time points.
The task is to predict whether the scenario is moving toward instability or remaining stable.
Core stability idea
Sepsis instability does not depend on one variable alone.
A patient may show an abnormal value and… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-sepsis-trajectory-instability-v0.1.clinical-five-node-sepsis-cascade-boundary-v0.7
What this repo does
This repository contains a Clarus v0.7 dataset modeling sepsis cascade boundary detection using a five-node cascade representation.
The dataset extends the earlier cascade-boundary structure by introducing uncertainty geometry.
The earlier question was:
Where is the cascade boundary?
v0.7 adds a second critical question:
How confident are we in that boundary and intervention judgment?
This allows Clarus to distinguish:
• confident boundary proximity• confident… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-five-node-sepsis-cascade-boundary-v0.7.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.3
Clinical Quad Infection Buffer Lag Coupling Sepsis Transition v1.3
Benchmark definition
Benchmark family: ClarusBenchmark layer: v1.3Geometry type: Failure Reconstruction GeometryDomain: Clinical stability systemsStructure: Quad coupling instability model
Primary question:
Can a model reconstruct the causal pathway that produced a failure state?
Evaluation requires identifying:
the ordered failure decision chain
the root policy error
the counterfactual recovery… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.3.clinical-recovery-window-sepsis-v1Clinical Recovery Window Sepsis Detection
Overview
This dataset tests whether a model can detect when a clinical system is still recoverable.
In severe infections such as sepsis, patients often pass through a finite period during which intervention can still reverse the trajectory toward collapse. This period is known as the recovery window.
Once the recovery window closes, physiological deterioration becomes increasingly difficult to reverse and the system moves toward irreversible failure.… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-recovery-window-sepsis-v1.clinical-control-sequence-sepsis-v1
Clinical Control Sequence Sepsis Detection
Overview
This dataset tests whether a model can detect whether a proposed intervention sequence is the correct stabilizing control sequence for a sepsis-like clinical system.
Complex systems are often not stabilized by a single action. They require the correct sequence of interventions delivered in the correct order as the system evolves.
The goal of this benchmark is to determine whether the control sequence meaningfully guides… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-control-sequence-sepsis-v1.clinical-control-horizon-sepsis-v1
Clinical Control Horizon Sepsis Detection
Overview
This dataset tests whether a model can detect whether a proposed control strategy has a sufficient stabilization horizon in a sepsis-like clinical system.
Some control strategies stabilize a system only briefly. Others maintain enough forward influence to guide the system into a durable recovery basin.
The goal of this benchmark is to determine whether the controller can reliably sustain stabilization across a meaningful… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-control-horizon-sepsis-v1.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.0
ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.0
What this repo does
This repository provides a Clarus v1.0 benchmark for sepsis transition under a four-variable clinical quad:
infection_load
buffer_capacity
lag_burden
coupling_stress
The v1.0 upgrade is Closed-Loop Control Geometry.
The task is no longer limited to detecting deterioration or ranking one intervention against another.
It tests whether a controller can:
choose the right path… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.0.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v0.6
What this repo does
This repository contains a Clarus v0.6 intervention pathway dataset focused on sepsis transition dynamics.
The dataset evaluates whether a model can determine if a proposed intervention meaningfully stabilizes a deteriorating septic system.
The task requires reasoning from:
system state
trajectory toward instability
boundary geometry
recovery geometry
intervention vector
projected trajectory consequence
The model cannot read the answer directly.
It must infer… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v0.6.clinical-recovery-stability-sepsis-v1Clinical Recovery Stability Sepsis Detection
Overview
This dataset tests whether a model can distinguish between temporary improvement and true structural recovery in a sepsis-like clinical system.
In many complex systems, short-term improvement can occur even while the system remains dangerously close to the instability boundary. Vital signs may improve temporarily, but the underlying dynamics may still favor relapse or collapse.
The task is therefore not simply detecting improvement, but… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-recovery-stability-sepsis-v1.clinical-instability-margin-sepsis-v1
Clinical Instability Margin Sepsis Detection
Overview
This dataset tests whether a model can detect when a clinical system is approaching the instability boundary.
In complex physiological systems, collapse rarely occurs suddenly. Instead the system gradually moves closer to a critical boundary where small disturbances can trigger rapid deterioration.
The goal is to determine whether the system is still safely inside the stability region or is approaching the instability… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-instability-margin-sepsis-v1.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v0.7
What this repo does
This repository contains a Clarus v0.7 dataset modeling sepsis transition using a quad-coupling system representation.
The dataset extends the v0.6 intervention layer by introducing uncertainty geometry.
The question addressed by earlier versions was:
Can the system be stabilized?
v0.7 adds a second critical question:
How confident are we in that conclusion?
This allows Clarus to distinguish three operational states:
• confident deterioration• confident… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v0.7.clinical-intervention-timing-sepsis-v1
Clinical Intervention Timing Sepsis Detection
Overview
This dataset tests whether a model can detect when an intervention is applied within the effective timing window of a sepsis-like clinical system.
In complex clinical systems, a correct intervention can still fail if it arrives too late relative to the system’s trajectory. The key question is not only whether an action is appropriate, but whether it occurs early enough to alter the path away from collapse.
The goal of… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-intervention-timing-sepsis-v1.clinical-sepsis-screening-antibiotic-escalation-coherence-risk-v0.1What this repo is for
Detect when sepsis signals
and screening plus antibiotic escalation
fall out of alignment
before
delayed treatment
and avoidable deterioration.
clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.1
Clinical Quad Infection Buffer Lag Coupling Sepsis Transition v1.1
What this repo does
This dataset evaluates whether a model can select the correct control policy when:
multiple sepsis interventions appear viable
early signals suggest improvement
alternative policies produce better long-term outcomes
The task is not prediction.
The task is selecting the correct action under uncertainty, feedback, and misleading signal structure.
Core quad
The system is… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.1.clinical-five-node-sepsis-cascade-boundary-v0.3
What this repo does
This dataset models sepsis cascade boundary approach using a Clarus five-node coupling framework combined with trajectory and system dynamics.
The goal is to predict whether a patient is approaching the sepsis cascade boundary.
The dataset introduces a dynamic forecasting layer that allows models to reason about motion through the stability manifold rather than relying only on static physiological snapshots.
Core five-node cascade… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-five-node-sepsis-cascade-boundary-v0.3.clinical-intervention-alignment-sepsis-v1Clinical Intervention Alignment Sepsis Detection
Overview
This dataset tests whether a model can determine whether a clinical intervention is aligned with the current system state.
In complex clinical systems such as sepsis, interventions do not have uniform effects. The same treatment may stabilize the system in one physiological state while having little effect—or even destabilizing the system—in another.
The benchmark evaluates whether models can detect when an intervention is structurally… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-intervention-alignment-sepsis-v1.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v0.8
What this repo does
This repository provides a Clarus v0.8 clinical quad dataset for detecting and reasoning about sepsis transition geometry.
The dataset models situations where a patient state is no longer contained within a single infection basin but is shifting between competing regimes such as:
infection burden with early inflammatory drift
septic shock transition
vasodilatory instability
multiorgan septic collapse
This is the conceptual upgrade introduced in Clarus v0.8.… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v0.8.clinical-stability-under-observation-sepsis-v1
Clinical Stability Under Observation Sepsis Detection
Overview
This dataset tests whether a model can detect whether a sepsis-like clinical system is intrinsically stable or only stable while under active observation and correction.
Some systems appear stable only because they are being continuously monitored and externally adjusted. If observation or correction intensity drops, the underlying dynamics can quickly pull the system back toward instability.
The goal of this… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-stability-under-observation-sepsis-v1.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.4
Clinical Quad Infection Buffer Lag Coupling Sepsis Transition v1.4
What this repo does
This repository contains a Clarus v1.4 benchmark dataset.
The v1.4 layer introduces Counterfactual Intervention Timing Geometry.
Earlier layers evaluate:
system state
trajectory
boundary proximity
recovery feasibility
intervention selection
control sequence correctness
temporal policy stability
failure reconstruction
v1.4 evaluates whether sepsis transition could still have been… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.4.clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.5Clinical Quad Oxygen Demand Buffer Lag Coupling Respiratory Collapse v1.5
What this repo does
This repository contains a Clarus v1.5 benchmark dataset.
The v1.5 layer introduces Counterfactual Rescue Path Sequencing Geometry.
Earlier Clarus layers evaluate:
• system state
• trajectory
• intervention selection
• control sequence correctness
• temporal policy stability
• failure reconstruction
• intervention timing
v1.5 evaluates a deeper reasoning task.
The benchmark asks whether a model can… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-quad-infection-buffer-lag-coupling-sepsis-transition-v1.5.clinical-reentry-instability-sepsis-v1
Clinical Reentry Instability Sepsis Detection
Overview
This dataset tests whether a model can detect whether a sepsis-like clinical system that appears to be recovering remains vulnerable to re-entry into instability.
Some systems move partially into recovery but continue to carry unresolved strain and attractive pull toward the instability basin. These systems can relapse under modest stress or incomplete stabilization.
The goal of this benchmark is to determine whether… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-reentry-instability-sepsis-v1.clinical-five-node-sepsis-cascade-boundary-v0.5
What this repo does
This repository provides a Clarus v0.5 cascade recovery geometry dataset modeling sepsis cascade transition with a five-node clinical structure.
Earlier Clarus datasets focused on state detection and boundary discovery.
Version v0.5 adds a recovery geometry layer that asks a stricter question:
Can the system still return to stability?
The task is binary classification over sepsis-linked deterioration states using:
• a five-node clinical cascade• trajectory… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-five-node-sepsis-cascade-boundary-v0.5.clinical-five-node-sepsis-cascade-boundary-v0.6
What this repo does
This repository contains a Clarus v0.6 intervention pathway dataset focused on sepsis cascade boundary dynamics.
The dataset evaluates whether a model can determine if a proposed intervention meaningfully stabilizes a deteriorating septic system represented as a five-node cascade.
The task requires reasoning from:
multi-node system state
trajectory toward instability
boundary geometry
recovery geometry
intervention vector
projected trajectory consequence
The… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-five-node-sepsis-cascade-boundary-v0.6.clinical-false-stability-sepsis-v1Clinical False Stability Sepsis Detection
Overview
This dataset tests whether a model can detect false stability in a clinical system.
False stability occurs when a system appears stable based on surface indicators, while deeper structural signals reveal that the system is already drifting toward collapse.
In real clinical environments this phenomenon appears frequently during severe infections such as sepsis. A patient's vital signs may temporarily stabilize even while the underlying… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-false-stability-sepsis-v1.clinical-latent-instability-sepsis-v1
Clinical Latent Instability Sepsis Detection
Overview
This dataset tests whether a model can detect latent instability in a sepsis-like clinical system.
Latent instability refers to structural conditions that make future destabilization likely even when overt instability is not yet visible. A system may appear broadly acceptable in the present while still carrying hidden load, high trigger sensitivity, and a directional pull toward later collapse.
The goal of this… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-latent-instability-sepsis-v1.clinical-compensation-collapse-sepsis-v1Clinical Compensation Collapse Sepsis Detection
Overview
This dataset tests whether a model can detect when a clinical system is moving from physiological compensation toward collapse.
In many critical illnesses such as sepsis, patients can appear stable for a period of time because the body temporarily compensates for rising stress. During this phase the system absorbs disturbances through physiological reserve.
However, compensation cannot continue indefinitely. Once reserve capacity is… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-compensation-collapse-sepsis-v1.clinical-hysteresis-sepsis-v1
Clinical Hysteresis Sepsis Detection
Overview
This dataset tests whether a model can detect hysteresis in a sepsis-like clinical system.
Hysteresis occurs when the future behavior of a system depends not only on its current state, but also on the path it took to get there. Two systems may appear similar at the present moment while having very different stability properties because one carries unresolved effects from prior stress.
The goal of this benchmark is to determine… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/clinical-hysteresis-sepsis-v1.ABX-CS-001_sepsis_treatment_response_decay-v0.1ABX-CS-001 Sepsis Treatment Response Decay
Purpose
Detect early failure of bloodstream infection clearing when clinical response stops tracking antibiotic exposure.
Core pattern
severity_index high
abx_exposure_index high
sepsis_coherence_index drops
cfu_persistence_vs_expected stays high
lactate_nonresponse_vs_expected stays high
later_septic_shock_or_persistent_bacteremia_flag appears
Files
data/train.csv
data/test.csv
scorer.py
Schema
Each row is one timepoint in a within series sepsis… See the full description on the dataset page: https://huggingface.co/datasets/ClarusC64/ABX-CS-001_sepsis_treatment_response_decay-v0.1.