INDEX

Explanations

proper nouns related to brands or organizations

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

Spatiallya listone of

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

IntoConstraints

-0.68

featureID

-0.55

TagMode

-0.55

:✨

-0.55

ModelAdmin

-0.53

$_"

-0.50

findpost

-0.48

expandindo

-0.48

webElement

-0.47

Datuak

-0.45

POSITIVE LOGITS

Etimología

0.41

 înd

0.41

 argint

0.41

EndProject

0.40

 plă

0.40

izarse

0.39

flatMap

0.39

 Projek

0.39

velas

0.38

 prosjek

0.38

Activations Density 0.011%

proper nouns related to brands or organizations

Spatiallya listone of

No Comments

No Known Activations