Dataset Viewer
Auto-converted to Parquet Duplicate
scenario_id
string
pressure
float64
drift_gradient
float64
feedback_gain
float64
correction_amplitude
float64
boundary_distance
float64
label_oscillatory_instability
int64
OE01
0.82
0.45
0.91
0.88
0.22
1
OE02
0.77
0.41
0.86
0.83
0.26
1
OE03
0.74
0.38
0.84
0.79
0.29
1
OE04
0.69
0.32
0.81
0.76
0.34
1
OE05
0.64
0.28
0.75
0.71
0.39
1
OE06
0.51
-0.12
0.41
0.36
0.71
0
OE07
0.47
-0.18
0.38
0.34
0.74
0
OE08
0.42
-0.22
0.35
0.31
0.79
0
OE09
0.38
-0.27
0.29
0.27
0.83
0
OE10
0.34
-0.31
0.24
0.22
0.87
0

OE — Oscillatory Instability Probe

This dataset is a periodic-table probe designed to detect oscillatory instability in complex systems.

Oscillatory instability occurs when corrective feedback overshoots system equilibrium, producing unstable cycles that amplify disturbances rather than dampening them.

The probe tests whether models can recognize the structural conditions that produce this behavior.


Geometry Symbol

OE

Oscillatory Instability

Family: Control Oscillation Failures


Mechanism

Oscillatory instability emerges when corrective actions interact with system pressure and feedback gain in a way that produces unstable oscillations.

Common features include:

  • strong feedback gain
  • large correction amplitude
  • positive drift toward instability
  • decreasing boundary distance

Signals

pressure
overall system stress

drift_gradient
direction of system trajectory

feedback_gain
strength of corrective response

correction_amplitude
magnitude of intervention adjustments

boundary_distance
distance to instability threshold


Prediction Target

label_oscillatory_instability

1 = unstable oscillatory regime forming
0 = stable or damped system behavior


Row Structure

Each row represents a system snapshot.

Signals are normalized between 0 and 1.

Higher values generally indicate stronger instability forces.


Files

data/train.csv
data/tester.csv
scorer.py


Evaluation

Predictions must contain:

scenario_id
prediction

Metrics reported:

accuracy
precision
recall
f1


Instability Geometry Probes

This dataset is part of a broader library of instability geometry probes:

CC — Cascade Collapse
BE — Buffer Exhaustion
DF — Delayed Feedback
OE — Oscillatory Instability

These probes collectively map fundamental instability patterns across complex systems.


License

MIT

Downloads last month
26