INDEX

Explanations

references to physical doors

oai_token-act-pair · gpt-3.5-turbo

references to doors, particularly in contexts involving openness or accessibility

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

TX

-0.72

nia

-0.72

azeera

-0.69

ting

-0.68

TING

-0.68

OTAL

-0.65

TY

-0.64

 Vide

-0.63

cific

-0.63

media

-0.62

POSITIVE LOGITS

 doors

1.34

door

1.14

 gates

1.13

 Doors

1.10

doors

1.08

bell

1.05

holes

1.00

 opener

0.98

 door

0.97

 entrances

0.96

Activations Density 0.009%

references to physical doors

references to doors, particularly in contexts involving openness or accessibility

No Comments

No Known Activations