INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 photographe

-0.42

 marié

-0.40

 soie

-0.39

 paneles

-0.39

 caisse

-0.38

 counsel

-0.38

 comedia

-0.37

 Nachrichten

-0.36

 indépendant

-0.36

 langues

-0.36

POSITIVE LOGITS

 tartalomajánló

0.63

AndroidJUnit

0.52

endphp

0.52

Autoritní

0.50

 <>",

0.49

URLException

0.49

yntaxException

0.49

IntoConstraints

0.49

BoxShadow

0.49

PostExecute

0.48

Activations Density 0.000%

No Known Activations

This feature has no known activations.