INDEX

Explanations

references to works of art

oai_token-act-pair · gpt-3.5-turbo

references to the concept of art

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

pora

-0.74

 Wyoming

-0.68

 Ingram

-0.66

ODE

-0.65

 Somalia

-0.64

VT

-0.63

 McGee

-0.62

 Crosby

-0.62

 Barker

-0.62

 Norway

-0.62

POSITIVE LOGITS

istry

1.50

emis

1.35

isans

1.34

works

1.29

ifice

1.22

esian

1.18

illery

1.11

ifacts

1.11

work

0.95

icles

0.91

Activations Density 0.021%

references to works of art

references to the concept of art

No Comments

No Known Activations