INDEX

Explanations

references to ethics in human or animal research

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

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/2-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.2.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

 president

-1.48

 vice

-1.45

 toast

-1.45

ials

-1.43

 spokesperson

-1.42

 executive

-1.40

mum

-1.39

 circulating

-1.39

 spokesman

-1.38

 boiler

-1.38

POSITIVE LOGITS

ocene

1.85

ledge

1.79

ĥ½

1.69

optera

1.52

alion

1.51

ureus

1.47

circuit

1.47

atorial

1.44

ailand

1.44

arynge

1.40

Activations Density 4.999%

references to ethics in human or animal research

No Comments

No Known Activations