INDEX

Explanations

various formatting and structural elements used in programming code

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

Angular Modules and headers

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enumii

-0.54

IsContent

-0.51

eqn

-0.51

 dAtA

-0.48

enumi

-0.47

 Volga

-0.45

Diwedd

-0.45

 pinulongan

-0.44

Коро

-0.44

 ModelExpression

-0.44

POSITIVE LOGITS

SharedModule

0.66

 CommonModule

0.45

 Already

0.41

 zawier

0.41

 AppModule

0.41

 NgModule

0.41

FormsModule

0.40

 already

0.39

PageModule

0.39

 BrowserModule

0.39

Activations Density 0.000%

various formatting and structural elements used in programming code

Angular Modules and headers

No Comments

No Known Activations