INDEX

Explanations

exactly the word "Bruce" with high activation

oai_token-act-pair · gpt-3.5-turbo

references to individuals named Bruce

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

esan

-0.73

FIX

-0.72

âĶĢ

-0.71

IX

-0.67

á

-0.66

isen

-0.66

 excise

-0.66

station

-0.65

mates

-0.65

 Erit

-0.63

POSITIVE LOGITS

 Bruce

3.51

Bruce

3.05

 Clint

1.45

 Batman

1.38

 Wayne

1.36

 Daryl

1.30

 Barbara

1.30

Batman

1.28

 Terry

1.22

 Barry

1.21

Activations Density 0.023%

exactly the word "Bruce" with high activation

references to individuals named Bruce

No Comments

No Known Activations