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

 cappello

-0.54

 forthwith

-0.48

 mute

-0.47

 Fremantle

-0.46

 clergy

-0.46

 sterilization

-0.45

 Genel

-0.45

 lyre

-0.45

 sterilized

-0.44

 diurnal

-0.44

POSITIVE LOGITS

 kasarigan

0.83

ValueStyle

0.82

 pinulongan

0.77

 autorytatywna

0.77

 EconPapers

0.73

 تضيفلها

0.71

 Мексичка

0.70

phazard

0.69

tagHelperRunner

0.69

 ویکی‌پدیا

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.