INDEX

Explanations

references to the PNG image format

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

PNG image files

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

webElementXpaths

-0.46

पया

-0.43

callee

-0.39

+#+

-0.39

spec

-0.38

didSet

-0.38

arif

-0.38

ecma

-0.37

type

-0.37

datatype

-0.36

POSITIVE LOGITS

png

0.74

 tartalomajánló

0.72

PNG

0.62

png

0.62

PNG

0.59

 pinulongan

0.58

 OnInit

0.57

 imageName

0.54

retudo

0.54

ngua

0.54

Activations Density 0.007%

references to the PNG image format

PNG image files

No Comments

No Known Activations