INDEX

Explanations

phrases related to controversial or dramatic situations involving people or events

oai_token-act-pair · gpt-3.5-turbo

references to political scrutiny and media perception

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

NetMessage

-0.75

NW

-0.71

è¦ļéĨĴ

-0.71

amacare

-0.70

rosso

-0.69

aukee

-0.68

leground

-0.67

foundation

-0.67

merce

-0.66

osures

-0.65

POSITIVE LOGITS

 afterwards

0.97

 later

0.89

 passers

0.88

 subsequent

0.83

 afterward

0.81

 thereafter

0.80

 subsequently

0.78

 himself

0.74

 onlook

0.74

 Later

0.73

Activations Density 1.178%

phrases related to controversial or dramatic situations involving people or events

references to political scrutiny and media perception

No Comments

No Known Activations