INDEX

Explanations

proper nouns related to the name "Scar"

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Scar" or "Scarborough."

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

cu

-0.81

 Dash

-0.81

 dash

-0.77

dash

-0.76

docker

-0.73

sonian

-0.73

nyder

-0.73

DF

-0.72

aqu

-0.71

DP

-0.69

POSITIVE LOGITS

 Scar

2.51

Scar

2.20

 Scarborough

2.12

 Priebus

1.55

 Rath

1.43

 wiret

1.33

 Bolton

1.23

Lis

1.00

Morning

0.96

 Thom

0.95

Activations Density 0.042%

proper nouns related to the name "Scar"

mentions of the name "Scar" or "Scarborough."

No Comments

No Known Activations