INDEX

Explanations

About Us, Q:

The neuron fires specifically on the word “formats.”

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

 this

-1.36

 allows

-1.28

 becomes

-1.24

an

-1.20

 makes

-1.20

 those

-1.18

 Those

-1.17

VolleyError

-1.16

 only

-1.16

all

-1.15

POSITIVE LOGITS

excellents

1.36

 ЛИ

1.31

 trám

1.27

Evet

1.26

 áng

1.23

 ДА

1.21

Daha

1.21

Nonetheless

1.21

of

1.19

ЛИ

1.19

Activations Density 0.000%