INDEX

Explanations

references to technical specifications or settings related to software or systems

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

ikipedia

-1.56

mph

-1.52

cox

-1.49

iella

-1.42

 mouths

-1.42

 cents

-1.40

apers

-1.40

 divor

-1.38

 econ

-1.36

aliana

-1.36

POSITIVE LOGITS

ĨĴ

1.86

 Programme

1.45

 vessel

1.43

²

1.42

Set

1.40

¯

1.39

++)

1.37

į

1.35

 programme

1.35

¦

1.35

Activations Density 5.574%

references to technical specifications or settings related to software or systems

No Comments

No Known Activations