INDEX

Explanations

phrases or sentences that include the word "who."

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

»

-2.05

ŀ

-1.94

¿½

-1.91

ķ

-1.75

 Clinic

-1.71

ľ

-1.62

Ľ

-1.60

 Breast

-1.58

¤

-1.57

Ĺ

-1.55

POSITIVE LOGITS

asting

1.75

 minded

1.52

 foul

1.48

ismiss

1.46

ushes

1.45

SEE

1.45

uttle

1.39

atter

1.39

asted

1.37

aven

1.37

Activations Density 3.078%

phrases or sentences that include the word "who."

No Comments

No Known Activations