INDEX

Explanations

describes qualifiers and emphasis

The neuron strongly activates on language describing travel or attraction reviews—especially mentions of scenic views and activities at a place.

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

if

-1.23

or

-1.12

 karena

-1.06

 млн

-1.03

 если

-1.02

 csak

-1.02

 takže

-1.02

 значит

-1.01

 gibi

-1.01

POSITIVE LOGITS

is

1.59

Especially

1.47

 especially

1.36

especially

1.29

are

1.24

 itself

1.22

 особенно

1.17

particularly

1.10

 Especially

1.10

 overall

1.09

Activations Density 0.028%