INDEX

Explanations

determines what something is or does

The neuron strongly activates on common structural function words (e.g. articles like “the,” copulas like “is,” and connectors like “where”) that organize sentence syntax.

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

 there

-1.75

 with

-1.63

 some

-1.58

 using

-1.53

 have

-1.33

 even

-1.28

and

-1.27

or

-1.20

 many

-1.18

at

-1.17

POSITIVE LOGITS

WillAppear

1.46

 crucial

1.41

 belangrijkste

1.28

 باید

1.25

antMatchers

1.23

用来

1.23

 pivotal

1.22

 مهم

1.19

 بايد

1.17

 besoins

1.15

Activations Density 0.074%