INDEX

Explanations

nothing

The neuron is primarily triggered by occurrences of the word “Nothing,” especially when it appears as the first token of a phrase.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 считают

0.63

 Новосибир

0.60

 एखा

0.59

ള്ള

0.59

вы

0.58

ACA

0.58

 somewhere

0.56

явля

0.56

כז

0.56

 synonym

0.55

POSITIVE LOGITS

 nada

1.15

 NOTHING

1.11

 nothing

1.10

 Nothing

1.10

nothing

1.02

 niets

0.93

Nothing

0.93

 Nada

0.91

nada

0.91

Nada

0.91

Activations Density 0.103%