INDEX

Explanations

words related to international contexts or interactions

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

Negative Logits

herty

-1.87

iful

-1.62

ening

-1.60

etts

-1.59

ering

-1.57

elt

-1.56

erman

-1.56

elta

-1.55

ellow

-1.54

enix

-1.53

POSITIVE LOGITS

 counsel

1.53

mediate

1.53

¦

1.45

Īĺ

1.43

stitial

1.42

Ń

1.42

±

1.40

 agencies

1.40

º

1.37

 agency

1.35

Activations Density 0.071%

words related to international contexts or interactions

No Known Activations