INDEX

Explanations

occurrences of marker tokens indicating the beginning of sections or paragraphs

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

the beginning of a paragraph or section in a document, indicating the start of a new topic or context

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

logout or exit

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

QMetaType

-0.70

 Wikimédia

-0.61

 actionMode

-0.57

 ContentValues

-0.56

numerusform

-0.55

viewDidLoad

-0.53

BeanUtils

-0.53

ChildScrollView

-0.52

 CommonModule

-0.51

.")

-0.51

POSITIVE LOGITS

logout

0.78

 logout

0.72

JMenu

0.64

 Logout

0.61

fromCharCode

0.59

LOGOUT

0.59

退出

0.58

logOut

0.56

Logout

0.55

 exit

0.54

Activations Density 0.025%

occurrences of marker tokens indicating the beginning of sections or paragraphs

the beginning of a paragraph or section in a document, indicating the start of a new topic or context

logout or exit

No Comments

No Known Activations