INDEX

Explanations

repeated or emphasized articles and words indicating location or quantity

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

area, water, top

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

FormTagHelper

-0.45

 propOrder

-0.45

}*/

-0.42

 AssemblyVersion

-0.42

esModule

-0.39

يميديا

-0.39

一说

-0.39

Cyfarwyddwr

-0.38

uação

-0.38

RTDA

-0.37

POSITIVE LOGITS

GEBURTSDATUM

0.56

 '\\;'

0.47

gdx

0.44

lgari

0.44

contentLoaded

0.44

 فريبيس

0.43

stead

0.42

 kohdetta

0.41

calaure

0.39

tences

0.39

Activations Density 0.002%

repeated or emphasized articles and words indicating location or quantity

area, water, top

No Comments

No Known Activations