INDEX

Explanations

words related to navigation or geographical terms

oai_token-act-pair · gpt-3.5-turbo

mentions of "Nav" or related terms associated with navigation

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

Reloaded

-0.70

 è£ı

-0.65

xual

-0.65

hift

-0.65

chnology

-0.64

 fertil

-0.64

FORE

-0.63

 spare

-0.63

 depos

-0.63

 Shattered

-0.62

POSITIVE LOGITS

arro

1.39

igator

1.38

ajo

1.29

igation

1.27

igating

1.25

igators

1.20

arre

1.11

istar

1.08

igate

1.08

arette

0.95

Activations Density 0.025%

words related to navigation or geographical terms

mentions of "Nav" or related terms associated with navigation

No Comments

No Known Activations