INDEX

Explanations

messages from senders

The neuron fires on occurrences of the word “message” (e.g. headings like “A Message” or “Message from …”).

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

-2.02

 capucha

-1.76

 kämpfen

-1.66

 اصلا

-1.62

 estos

-1.61

 چیه

-1.61

۱۱

-1.57

 praises

-1.55

 использовании

-1.53

 когато

-1.48

POSITIVE LOGITS

to

1.81

as

1.70

“

1.49

message

1.43

 indicate

1.42

 concise

1.34

Our

1.33

 provide

1.30

 adhere

1.30

 they

1.27

Activations Density 0.016%