INDEX

Explanations

words related to occupation and professional roles

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/1-res-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hip

-1.80

 terror

-1.67

 face

-1.55

htra

-1.51

 figure

-1.51

hes

-1.44

nol

-1.43

 heed

-1.41

usterity

-1.39

WHM

-1.39

POSITIVE LOGITS

art

1.61

Ĵ

1.56

ilic

1.56

 polls

1.44

orate

1.42

orem

1.41

ong

1.41

iles

1.40

 tast

1.39

rant

1.37

Activations Density 0.056%

words related to occupation and professional roles

No Comments

No Known Activations