INDEX

Explanations

specified, varying, determined

The main thing this neuron does is find numeric values (tokens representing numbers or measurements).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

it

-1.59

of

-1.43

 there

-1.38

all

-1.23

 with

-1.20

 having

-1.20

now

-1.19

 this

-1.16

 first

-1.16

out

-1.14

POSITIVE LOGITS

 determinado

1.41

 diversos

1.30

 specified

1.29

 различ

1.28

 tertentu

1.27

 varies

1.27

 varying

1.24

 selected

1.23

一定

1.22

 quelconque

1.22

Activations Density 0.134%