INDEX

Explanations

action verbs followed by articles/prepositions

The neuron fires on prominent action‐oriented verbs (often imperatives) used to introduce tasks or instructions.

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

͌

-1.22

)):

-1.20

 Tatsache

-1.18

 Pläne

-1.17

Hilsen

-1.10

 leichten

-1.09

 Пусть

-1.09

 Gründe

-1.09

 innych

-1.06

$-$

-1.05

POSITIVE LOGITS

 these

2.17

 this

1.70

the

1.53

and

1.48

 each

1.37

这些

1.24

all

1.17

该

1.12

 این

1.12

 statut

1.11

Activations Density 0.052%