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

=’

-0.47

fromCharCode

-0.44

']=='

-0.41

 politically

-0.40

Dij

-0.39

ectl

-0.38

JsonPropertyName

-0.38

ectoria

-0.38

 boxes

-0.37

 '\\;'

-0.37

POSITIVE LOGITS

ViewImports

0.54

BufferException

0.49

nasel

0.48

 يتيمه

0.45

 nakalista

0.43

AndEndTag

0.42

 ressemb

0.42

-------------</

0.41

 Infór

0.41

+#+#

0.40

Activations Density 0.000%

No Known Activations

This feature has no known activations.