INDEX

Explanations

Additionally leads to other

The neuron detects additive transition phrases (e.g. “In addition,” “Additionally,” etc.).

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

 that

-1.55

vidades

-1.38

 from

-1.35

\}$,

-1.30

 głównie

-1.25

]}{

-1.25

Մ

-1.24

 which

-1.24

}]

-1.23

איך

-1.21

POSITIVE LOGITS

2.23

 there

1.67

 otros

1.59

other

1.54

 também

1.52

 outras

1.43

 هناك

1.41

 other

1.39

you

1.35

 autre

1.34

Activations Density 0.017%