INDEX

Explanations

not possible

The neuron flags words and phrases expressing negation or impossibility (e.g. “not,” “cannot,” “impossible,” “can’t”).

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

 plutôt

-1.24

 stärker

-1.11

 fiecare

-1.10

INGE

-1.08

 vähän

-1.08

fäl

-1.07

 délais

-1.06

 reager

-1.06

mäß

-1.02

 vêtements

-1.02

POSITIVE LOGITS

any

2.25

 unless

2.22

 because

1.63

 anymore

1.61

 without

1.56

 إلا

1.51

 this

1.43

 Unless

1.39

 with

1.38

有任何

1.38

Activations Density 0.092%