INDEX

Explanations

names of people or locations likely related to news or political scandals

oai_token-act-pair · gpt-3.5-turbo

proper nouns related to individuals and significant events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

bye

-0.80

esome

-0.70

erness

-0.69

theless

-0.68

ed

-0.67

akening

-0.66

ONE

-0.64

 teeth

-0.64

bon

-0.63

 Dill

-0.63

POSITIVE LOGITS

 Abram

0.95

agall

0.80

urations

0.76

adan

0.75

coni

0.75

ashtra

0.74

advert

0.73

asta

0.71

merce

0.71

ventions

0.70

Activations Density 0.027%

names of people or locations likely related to news or political scandals

proper nouns related to individuals and significant events

No Comments

No Known Activations