INDEX

Explanations

information related to the concept of exposure, in various contexts such as health, finance, and news dissemination

oai_token-act-pair · gpt-3.5-turbo

mentions of exposure, particularly in a health or safety context

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

 Mattis

-0.69

 Frey

-0.68

iddles

-0.68

bane

-0.67

igh

-0.66

yy

-0.66

Econom

-0.64

 Polit

-0.64

bard

-0.63

:{

-0.63

POSITIVE LOGITS

 exposure

3.78

 exposures

2.97

 Exposure

2.86

posure

2.30

 exposed

1.65

 exposing

1.40

 expose

1.39

 visibility

1.36

 susceptibility

1.33

 immersion

1.29

Activations Density 0.023%

information related to the concept of exposure, in various contexts such as health, finance, and news dissemination

mentions of exposure, particularly in a health or safety context

No Comments

No Known Activations