INDEX

Explanations

entities related to locations and organizations

oai_token-act-pair · gpt-3.5-turbo

names of places and organizations, along with associated numerical or categorized data

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

 ..........

-0.45

tap

-0.45

Enlarge

-0.43

audi

-0.42

endi

-0.41

odynam

-0.40

Copyright

-0.40

 mids

-0.40

Thumbnail

-0.40

alyst

-0.40

POSITIVE LOGITS

hower

0.63

determination

0.51

 Hera

0.50

ItemThumbnailImage

0.50

ebted

0.43

Ã±a

0.42

schild

0.42

ahime

0.42

haar

0.41

holy

0.41

Activations Density 6.367%

entities related to locations and organizations

names of places and organizations, along with associated numerical or categorized data

No Comments

No Known Activations