INDEX

Explanations

mentions of affiliate links and marketing-related terms

oai_token-act-pair · gpt-3.5-turbo

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

senal

-0.78

OPLE

-0.68

 Kinnikuman

-0.67

GY

-0.65

MQ

-0.65

çķ

-0.64

 subtract

-0.63

åĤ

-0.63

===

-0.63

 externalToEVAOnly

-0.63

POSITIVE LOGITS

idav

1.44

licted

1.44

irm

1.44

inity

1.43

irmed

1.43

ordable

1.42

irmation

1.37

luent

1.36

irming

1.34

iliate

1.31

Activations Density 0.012%

mentions of affiliate links and marketing-related terms

No Comments

No Known Activations