INDEX

Explanations

references to the Oval Office

oai_token-act-pair · gpt-3.5-turbo

references to the Oval Office

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

SOURCE

-0.73

unks

-0.72

TOR

-0.69

ARP

-0.69

weeney

-0.67

NRS

-0.67

uther

-0.66

 delinquent

-0.66

Domin

-0.65

ulia

-0.64

POSITIVE LOGITS

 Oval

1.06

 Office

0.78

croft

0.77

lace

0.75

 Medal

0.74

flower

0.73

OTUS

0.72

minster

0.72

bowl

0.70

 Heights

0.70

Activations Density 0.007%

references to the Oval Office

references to the Oval Office

No Comments

No Known Activations