INDEX

Explanations

mentions of locations within political contexts, specifically focusing on the Oval Office

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Oval Office."

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

uther

-0.76

ARP

-0.75

SOURCE

-0.72

weeney

-0.69

ictions

-0.68

HCR

-0.66

arg

-0.63

 delinquent

-0.63

unks

-0.63

TOR

-0.63

POSITIVE LOGITS

 Oval

1.15

OTUS

0.82

 Office

0.79

flower

0.72

bowl

0.72

croft

0.71

minster

0.69

 Oath

0.69

rium

0.67

Ħ¢

0.67

Activations Density 0.006%

mentions of locations within political contexts, specifically focusing on the Oval Office

occurrences of the term "Oval Office."

No Comments

No Known Activations