INDEX

Explanations

references to specific instances or illustrations

oai_token-act-pair · gpt-3.5-turbo

references to illustrative instances or cases

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

yss

-0.88

usalem

-0.74

 livest

-0.73

olulu

-0.71

 satell

-0.71

uld

-0.69

IER

-0.65

ternity

-0.65

 irrad

-0.64

 subscription

-0.64

POSITIVE LOGITS

 examples

1.24

amples

0.99

 Examples

0.99

uations

0.89

baugh

0.86

example

0.86

DragonMagazine

0.85

Examples

0.81

 illustrating

0.79

 attRot

0.78

Activations Density 0.011%

references to specific instances or illustrations

references to illustrative instances or cases

No Comments

No Known Activations