INDEX

Explanations

references to default system settings in code or software configurations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

ChildScrollView

-0.66

 PeEnEo

-0.59

featureID

-0.56

 ब्रेकडाउन

-0.54

SourceChecksum

-0.53

 nonUne

-0.52

 iconLine

-0.52

SharedCtor

-0.51

Tikang

-0.48

RenderAtEndOf

-0.47

POSITIVE LOGITS

styleType

0.59

 conformal

0.46

UseVisualStyle

0.43

mobileqq

0.42

 JADX

0.40

 getSystem

0.40

 standard

0.39

 voices

0.38

 carne

0.38

standard

0.37

Activations Density 0.003%

references to default system settings in code or software configurations

No Comments

No Known Activations