INDEX

Explanations

mentions of personal preferences or tastes

oai_token-act-pair · gpt-3.5-turbo

phrases related to personal tastes and preferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Chern

-0.81

BILITY

-0.80

boxing

-0.79

ICA

-0.78

Assembly

-0.75

 Brotherhood

-0.71

Ded

-0.69

ANCE

-0.69

Veter

-0.68

à¤

-0.68

POSITIVE LOGITS

omething

1.24

avorite

1.17

ettings

1.14

ometimes

1.10

cape

0.99

hops

0.98

uggest

0.97

 tastes

0.94

ystem

0.93

wana

0.93

Activations Density 0.019%

mentions of personal preferences or tastes

phrases related to personal tastes and preferences

No Comments

No Known Activations