INDEX

Explanations

dates written in abbreviated form with 'd'

oai_token-act-pair · gpt-3.5-turbo

repeated instances of the letter 'd'

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 constitu

-0.66

 Remem

-0.65

Nun

-0.63

 Corvette

-0.60

 salv

-0.58

WAYS

-0.58

 suscept

-0.58

 folds

-0.57

 Leone

-0.57

OIL

-0.57

POSITIVE LOGITS

ividual

1.42

ynamic

1.38

aniel

1.27

iamond

1.24

iary

1.21

elta

1.20

ynam

1.20

owntown

1.16

iscover

1.16

orf

1.15

Activations Density 0.069%

dates written in abbreviated form with 'd'

repeated instances of the letter 'd'

No Comments

No Known Activations