INDEX

Explanations

mentions of the word "Az" followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to the name "Az" along with its variations in context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

aceous

-0.75

 Fargo

-0.74

ACTED

-0.74

dfx

-0.69

 Norn

-0.69

LESS

-0.68

 Sporting

-0.64

mercial

-0.63

 ease

-0.63

 Copenhagen

-0.62

POSITIVE LOGITS

azel

1.18

tec

1.15

eri

1.06

uma

0.95

rael

0.92

er

0.90

orius

0.89

anas

0.88

eta

0.87

irez

0.87

Activations Density 0.023%

mentions of the word "Az" followed by a number

references to the name "Az" along with its variations in context

No Comments

No Known Activations