INDEX

Explanations

mentions or references to the character "Wonder Woman"

oai_token-act-pair · gpt-3.5-turbo

references to the character "Wonder Woman."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Mald

-0.69

 retracted

-0.67

axis

-0.66

lic

-0.66

itual

-0.65

Ã¤

-0.65

 committed

-0.64

 extortion

-0.64

 transient

-0.63

fee

-0.63

POSITIVE LOGITS

 Wonder

3.86

Wonder

2.73

 Wonders

2.29

 Wond

1.52

 wonder

1.28

 Superman

1.25

 Diana

1.17

 Wizard

1.16

 Patty

1.15

 Wonderland

1.15

Activations Density 0.015%

mentions or references to the character "Wonder Woman"

references to the character "Wonder Woman."

No Comments

No Known Activations