INDEX

Explanations

No Explanations Found

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

IX

-1.56

completely

-1.53

»

-1.52

Into

-1.47

edly

-1.37

Advertisement

-1.37

 densely

-1.34

npmjs

-1.33

 vigorous

-1.33

Holy

-1.33

POSITIVE LOGITS

upal

1.72

uable

1.64

estic

1.62

uen

1.58

asons

1.56

cester

1.55

enne

1.53

amps

1.53

 pictured

1.51

ided

1.50

Activations Density 0.000%

No Known Activations

This feature has no known activations.