INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

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

 sillas

-0.54

 pén

-0.47

spreis

-0.47

 translateY

-0.46

TabIndex

-0.46

ollectionView

-0.46

bootstrapcdn

-0.45

 prestation

-0.42

 recevrez

-0.42

VIAF

-0.42

POSITIVE LOGITS

migrationBuilder

0.64

")[

0.61

enumii

0.60

يميديا

0.57

ंदीखरीदारी

0.56

MethodManager

0.56

contentLoaded

0.55

 gekennzeichnet

0.55

 safe

0.54

 <>",

0.54

Activations Density 0.000%

No Known Activations

This feature has no known activations.