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

 للمعارف

-0.92

 propOrder

-0.91

 незавершена

-0.90

 IBOutlet

-0.83

 Wikimedijinoj

-0.82

RTLI

-0.76

 дописавши

-0.75

()]);

-0.74

AnchorStyles

-0.72

Datuak

-0.71

POSITIVE LOGITS

turn

0.43

 înainte

0.43

 quantités

0.41

îl

0.40

 altså

0.39

共に

0.39

 côtes

0.38

 poches

0.38

 carattere

0.37

 scienza

0.37

Activations Density 0.000%

No Known Activations

This feature has no known activations.