INDEX

Explanations

history

The neuron activates strongly on terms referring to logs or histories (e.g. “call log,” “history,” “Logs,” “History”).

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

↵↵

-1.59

饜

-1.58

are

-1.53

 Political

-1.45

 ſta

-1.43

-1.30

१

-1.27

-1.25

<sup>

-1.24

she

-1.23

POSITIVE LOGITS

Välkommen

1.60

点此

1.48

of

1.44

᪥

1.43

厹

1.42

AtA

1.42

 nedeniyle

1.38

媖

1.34

Väl

1.34

Yuki

1.32

Activations Density 0.061%