Datasets:

ClarusC64
/

f1-latent-cross-coupling-thermal-load-instability-v0.1

observable_state string	latent_instability_score float64	cross_coupling_intensity float64	hidden_state_index float64	activation_threshold_distance float64	tyre_temp_load float64	brake_temp_load float64	power_unit_heat_load float64	cooling_efficiency float64	stabilization_buffer int64	label_thermal_load_instability float64
stable	0.84	0.81	0.85	0.18	0.88	0.82	0.8	0.32	1	null
stable	0.76	0.72	0.78	0.26	0.83	0.78	0.75	0.38	1	null
managed	0.52	0.49	0.53	0.55	0.7	0.58	0.57	0.6	0	null
cooling-stable	0.3	0.31	0.29	0.8	0.36	0.35	0.34	0.72	0	null
managed	0.69	0.65	0.7	0.33	0.81	0.73	0.71	0.42	1	null
balanced	0.57	0.59	0.58	0.4	0.63	0.61	0.6	0.58	0	null
stable	0.87	0.84	0.86	0.15	0.91	0.85	0.83	0.28	1	null
cooling-stable	0.34	0.36	0.33	0.76	0.39	0.38	0.37	0.68	0	null
managed	0.61	0.58	0.63	0.37	0.72	0.69	0.68	0.5	0	null
stable	0.9	0.83	0.88	0.12	0.93	0.86	0.85	0.26	1	null

What this repo does

Detects hidden thermal instability before performance loss appears.

Focus: temperature-driven failure across interacting systems.

label_thermal_load_instability

1 → thermal regime will force performance drop
0 → thermal state remains stable

Thermal failure is rarely single-source.

It emerges from interaction:

label = 1 if:

Performance loss is often thermal before it is visible.

This dataset detects:

Primary: missed_latent_activation_rate

Secondary: false_activation_rate

This sits in the latent detection layer.

It detects hidden thermal instability before performance collapse appears.

Used for:

Suitable for:

MIT