INDEX

Explanations

outgoing [thing sent]

The neuron primarily activates on the word “outgoing.”

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

of

-2.30

 reempla

-2.20

跺

-2.17

𝟭

-2.17

 sorprender

-2.11

-2.08

 появится

-2.05

 negar

-2.03

稣

-2.03

 olvid

-2.02

POSITIVE LOGITS

ścić

2.78

～～

2.73

queles

2.58

刧

2.58



2.58

 necessaria

2.55

The

2.50

部分

2.48



2.39

ated

2.36

Activations Density 0.005%