INDEX

Explanations

depending on

The neuron fires strongly on the word “Depending” (and the following “on”)—i.e. it detects “Depending on …” conditional phrases.

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

 semua

0.71

 viral

0.71

 explosion

0.71

 pharmaceutical

0.70

 virus

0.68

ٹنے

0.68

 helpline

0.67

 Pharmaceutical

0.66

 افز

0.66

バイル

0.65

POSITIVE LOGITS

Depending

1.01

depending

0.93

 depending

0.84

degrees

0.80

↵

0.75

你是

0.75

情况

0.74

degree

0.74

ри

0.72

 Depending

0.72

Activations Density 0.062%