INDEX

Explanations

occurrences of the word "wrote."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

½

-2.10

ı

-1.90

·¸

-1.84

µ

-1.61

»

-1.60

 extent

-1.54

 Malaysia

-1.53

Ŀ

-1.52

¼

-1.51

 granted

-1.48

POSITIVE LOGITS

 about

2.07

about

1.99

 About

1.70

 down

1.65

imental

1.56

cknowled

1.54

nail

1.51

tet

1.48

initis

1.48

iously

1.47

Activations Density 0.080%

occurrences of the word "wrote."

No Comments

No Known Activations