INDEX

Explanations

the beginning of structured data or text inputs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.21.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

 entendu

-0.30

运

-0.30

 Viel

-0.28

SPO

-0.28

CIÓN

-0.27

itu

-0.26

 hats

-0.25

 staff

-0.25

 تك

-0.25

brado

-0.25

POSITIVE LOGITS

 autorytatywna

0.84

NameInMap

0.83

 Administrativna

0.78

expandindo

0.77

IUrlHelper

0.76

OGND

0.73

adaptiveStyles

0.71

 nahilalakip

0.70

 resourceCulture

0.68

 CreateTagHelper

0.68

Activations Density 0.120%

the beginning of structured data or text inputs

No Comments

No Known Activations