INDEX

Explanations

actual vs potential

The neuron identifies abstract nouns that denote states or measures of capacity (e.g. actual vs. potential).

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

dulo

-1.37

溲

-1.33

}),

-1.29

秕

-1.27

higen

-1.27

 alimenti

-1.25

 système

-1.24



-1.23

 convain

-1.23

 可愛い

-1.23

POSITIVE LOGITS

and

1.69

 because

1.38

 until

1.36

in

1.32

</strong>

1.29

 though

1.22

 where

1.19

 gedrag

1.17

囖

1.16

 here

1.16

Activations Density 0.050%