INDEX

Explanations

highly emotive expressions or phrases associated with strong feelings

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__mlp-sm_processed/0-mlp-sm

Prompts (Dashboard)

32,768 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

torch.float32

Hook Name

blocks.0.hook_mlp_out

Hook Layer

Architecture

standard

Context Size

128

Dataset

EleutherAI/the_pile_deduplicated

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ols

-1.79

hene

-1.56

fulness

-1.50

hol

-1.47

hov

-1.46

hors

-1.43

mong

-1.42

 agar

-1.39

ructose

-1.39

ethanol

-1.36

POSITIVE LOGITS

selector

1.52

 rights

1.49

ly

1.44

 Specifically

1.43

 Plaintiff

1.39

))?

1.39

sequently

1.39

itoneal

1.39

istics

1.38

acent

1.35

Activations Density 0.021%

highly emotive expressions or phrases associated with strong feelings

No Known Activations