INDEX

Explanations

instances of content deletion or content that is unreadable

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

code structure or placeholders

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

__":

-0.40

 réalis

-0.39

InputBorder

-0.36

[]):

-0.35

JpaRepository

-0.35

borderFill

-0.34

DockStyle

-0.34

RectangleBorder

-0.34

”:

-0.33

 guère

-0.33

POSITIVE LOGITS

帖最后由

0.79

httphttps

0.71

endregion

0.61

0.57

AsUp

0.56

 withheld

0.49

Derbyniad

0.49

ViewFeatures

0.48

[*]

0.48

ंदीखरीदारी

0.46

Activations Density 0.019%

instances of content deletion or content that is unreadable

code structure or placeholders

No Comments

No Known Activations