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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 createState

-0.60

LookAnd

-0.52

 Inscrivez

-0.50

 aikana

-0.49

 kiin

-0.48

 citoy

-0.48

drawSprites

-0.47

::_('

-0.45

 héroe

-0.44

 héros

-0.43

POSITIVE LOGITS

 autorytatywna

0.90

CodeAttribute

0.84

ValueStyle

0.82

SuppressMessage

0.79

Attribute

0.72

 nahilalakip

0.64

 ***!

0.60

basicConfig

0.56

NameInMap

0.54

 Attribute

0.53

Activations Density 0.000%

No Known Activations

This feature has no known activations.