INDEX

Explanations

other followed by classification

The neuron activates primarily on the word “other.”

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

is

-2.23

it

-2.19

 here

-1.91

 place

-1.72

This

-1.72

 especially

-1.66

 where

-1.66

 there

-1.65

to

-1.64

 when

-1.63

POSITIVE LOGITS

 berge

1.43

Бли

1.42

 أيض

1.40

 الذات

1.38

不错

1.38

 rése

1.38

 kupa

1.38

 növény

1.38

ギュア

1.38

 kuin

1.38

Activations Density 0.031%