INDEX

Explanations

programming-related syntax elements or structures

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

C++ preprocessor directives

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

bootstrapcdn

-0.54

 تكبرها

-0.52

IndentedString

-0.50

 мәкал

-0.47

 onPostExecute

-0.44

 kaarangay

-0.43

RetentionPolicy

-0.43

AccessorTable

-0.42

BitmapFactory

-0.42

 Italijani

-0.41

POSITIVE LOGITS

الحياه

0.54

and

0.50

verifyException

0.50

 AssemblyCulture

0.50

principalTable

0.47

LinkId

0.41

 Ayers

0.40

 nahilalakip

0.39

 Potts

0.39

mentado

0.37

Activations Density 0.057%

programming-related syntax elements or structures

C++ preprocessor directives

No Comments

No Known Activations