INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_1m/average_l0_107

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

1,048,576

Data Type

float32

Hook Name

blocks.12.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

ÚBL

-0.54

الإنجليزية

-0.52

emia

-0.49

 wikihow

-0.49

vieve

-0.49

itecture

-0.47

attva

-0.47

gemeen

-0.47

detta

-0.47

>",

-0.46

POSITIVE LOGITS

AndEndTag

0.81

Portale

0.74

setVerticalGroup

0.65

 незавершена

0.62

 parlato

0.61

 مشين

0.61

complexContent

0.60

 propOrder

0.60

IBOutlet

0.59

KURZBESCHREIBUNG

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.