INDEX

Explanations

specifications related to graphical elements, particularly dimensions and formats

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

width attribute in markup

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

 also

-0.32

 Humanitarian

-0.29

ForRow

-0.29

ReusableCell

-0.28

AddTagHelper

-0.28

OfDay

-0.27

 does

-0.27

).

-0.27

homonymie

-0.26

 tambien

-0.26

POSITIVE LOGITS

 estekak

0.66

postsleuth

0.64

 defaultstate

0.63

DeleteBehavior

0.63

 Taktlose

0.63

 ویکی‌پدی

0.63

Jeografia

0.62

 Италијани

0.60

 AssemblyCulture

0.60

 Савезне

0.59

Activations Density 0.003%

specifications related to graphical elements, particularly dimensions and formats

width attribute in markup

No Comments

No Known Activations