INDEX

Explanations

phrases related to additional information or supplementary content

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

Ĵ

-2.40

Ļª

-2.31

Ĥ¬

-2.28

¼

-2.23

ĨĴ

-2.17

½

-1.93

Ģ

-1.86

Į

-1.85

Ļ

-1.83

º

-1.82

POSITIVE LOGITS

 finding

1.50

edia

1.46

enberg

1.45

 suppose

1.39

idegger

1.31

ype

1.30

ferential

1.27

yler

1.25

 black

1.24

coming

1.22

Activations Density 0.083%

phrases related to additional information or supplementary content

No Comments

No Known Activations