INDEX

Explanations

references to notable or widely recognized individuals and entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Configuration

ctigges/pythia-70m-deduped__res-sm_processed/1-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.1.hook_resid_post

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

aned

-1.96

rote

-1.55

arter

-1.54

ilateral

-1.46

UID

-1.46

romycin

-1.44

oda

-1.42

anos

-1.41

orb

-1.39

oner

-1.39

POSITIVE LOGITS

Ļª

1.98

áĢº

1.88

 footsteps

1.76

 fame

1.74

ĻĤ

1.73

ness

1.66

 names

1.65

mong

1.64

 moments

1.62

 patrons

1.59

Activations Density 0.158%

references to notable or widely recognized individuals and entities

No Comments

No Known Activations