INDEX

Explanations

directional geographical locations

The neuron responds to directional or positional terms—especially compass points (north, south, east, west) and related ordinal/location words (first, second, end).

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

on

-1.91

of

-1.66

 because

-1.36

 about

-1.35

 from

-1.31

 every

-1.29

The

-1.28

if

-1.27

in

-1.26

↵↵↵

-1.21

POSITIVE LOGITS

一个

1.52

ynchronously

1.41

ーーー

1.38

 esponja

1.32

 اوم

1.31

 rotación

1.31

品种

1.30

Примі

1.30

 voiture

1.30

自分の

1.29

Activations Density 0.049%