INDEX

Explanations

phrases related to confidentiality and secrets

oai_token-act-pair · gpt-3.5-turbo

terms and phrases associated with confidentiality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

stad

-0.66

rep

-0.66

dies

-0.64

omy

-0.63

ost

-0.62

(#

-0.61

 feat

-0.61

saw

-0.61

itching

-0.61

 Conquer

-0.61

POSITIVE LOGITS

 confidential

3.81

 confidentiality

2.15

 anonymous

1.39

 redacted

1.36

idential

1.35

 informant

1.29

 informants

1.28

 sensitive

1.27

 secret

1.26

 anonymously

1.20

Activations Density 0.021%

phrases related to confidentiality and secrets

terms and phrases associated with confidentiality

No Comments

No Known Activations