INDEX

Explanations

existence of items or features

This neuron activates strongly on very common English function words (e.g. articles and auxiliaries like “the,” “have/has,” “or”) and punctuation (commas).

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

|]

-1.42

 there

-1.34

even

-1.16

()

-1.16

 that

-1.16

you

-1.14

-1.13

 noin

-1.12

 stiker

-1.12

 nawet

-1.12

POSITIVE LOGITS

的公司

1.49

 dní

1.44

 男人

1.33

1.31

Особенности

1.28

 svoje

1.27

 était

1.27

 include

1.26

 tristes

1.22

莫

1.19

Activations Density 0.083%