INDEX

Explanations

instances of the underscore character, often indicating placeholders or hidden elements in code

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_4/width_65k/average_l0_66

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.4.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

AddTagHelper

-0.62

DrawerToggle

-0.49

ID

-0.45

Fig

-0.45

Fig

-0.45

MediaStore

-0.42

ImageContext

-0.41

FileReader

-0.40

comando

-0.40

MessageTagHelper

-0.40

POSITIVE LOGITS

 للمعارف

0.89

رشف

0.86

TagMode

0.81

 whoſe

0.79

 ſtate

0.75

 ſtre

0.71

ſy

0.70

rawDesc

0.70

 purpoſe

0.69

GEBURTSDATUM

0.68

Activations Density 0.008%

instances of the underscore character, often indicating placeholders or hidden elements in code

No Comments

No Known Activations