INDEX

Explanations

fire and flammable risks

This neuron activates on tokens in formal informational or news‐style text—especially proper names, titles, and other metadata (e.g. interview subjects, organizations, and headings).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

SqlServer

0.63

 đá

0.57

Sql

0.54

Tic

0.54

Sql

0.53

໋

0.53

 दांत

0.51

ワク

0.50

 روح

0.49

reonine

0.49

POSITIVE LOGITS

 fire

2.81

 firefighting

2.69

 firefighters

2.61

 Fire

2.55

Fire

2.50

 firefighter

2.50

 flames

2.45

火

2.45

 fires

2.44

 آتش

2.42

Activations Density 0.276%