INDEX

Explanations

emergency situations and declarations

The neuron activates on tokens that signal official emergency or crisis declarations and associated conditions (e.g., “state of emergency,” “public health emergency,” “Level 2 Drought Condition,” “major incident declared”).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 które

-1.58

 whom

-1.57

 occuper

-1.51

 rempliss

-1.47

 เมื่อ

-1.45

蛱

-1.44

tinh

-1.44

chose

-1.43

borsa

-1.43

 nossos

-1.42

POSITIVE LOGITS

in

2.20

1.91

for

1.85

if

1.81

at

1.79

 there

1.72

 just

1.72

بيه

1.60

not

1.59

1.58

Activations Density 0.021%