INDEX

Explanations

code-related elements, specifically focusing on URLs and objects in programming contexts

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

to sanitize, need to update, piece of, data contains

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

setAuto

-0.31

ter

-0.30

 auto

-0.30

et

-0.29

 past

-0.27

gam

-0.27

-0.26

ber

-0.26

al

-0.26

POSITIVE LOGITS

UnusedPrivate

0.95

 ujednoznacz

0.79

 متعلقه

0.79

ReusableCell

0.78

 ivelany

0.75

findpost

0.72

InjectAttribute

0.71

ItemBackground

0.71

 ostavi

0.70

WebElementEntity

0.70

Activations Density 0.351%

code-related elements, specifically focusing on URLs and objects in programming contexts

to sanitize, need to update, piece of, data contains

No Comments

No Known Activations