INDEX

Explanations

the name "Francesco" or variations of it

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Frances"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ocrine

-0.83

eering

-0.69

ansas

-0.66

eers

-0.66

 Helsinki

-0.65

pocket

-0.63

eer

-0.63

 sockets

-0.63

uyomi

-0.63

Seattle

-0.61

POSITIVE LOGITS

terness

0.93

olini

0.92

 Frances

0.88

chini

0.87

cus

0.82

ca

0.79

atell

0.78

ella

0.78

utical

0.77

etta

0.76

Activations Density 0.004%

the name "Francesco" or variations of it

mentions of the name "Frances"

No Comments

No Known Activations