INDEX

Explanations

phrases related to individuality and self-expression

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 34-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_34/width_16k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.34.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

 lamang

-0.68

 bowiem

-0.59

十分的

-0.50

 zimowa

-0.49

iż

-0.48

 dourado

-0.48

 poichè

-0.48

 doskon

-0.46

 mnoho

-0.46

 sahiptir

-0.46

POSITIVE LOGITS

 guys

0.66

 dude

0.66

 fuckin

0.63

 yeah

0.58

 chill

0.58

 kinda

0.57

 dunno

0.57

 stuff

0.56

 dupe

0.55

 shitty

0.54

Activations Density 0.093%

phrases related to individuality and self-expression

No Comments

No Known Activations