INDEX

Explanations

financial figures or news headlines related to companies and individuals

oai_token-act-pair · gpt-3.5-turbo

occurrences of slashes or division symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 accus

-0.74

 stalking

-0.74

 assemb

-0.73

 inqu

-0.72

 record

-0.72

 imperson

-0.70

 reve

-0.70

 demol

-0.69

 explan

-0.68

 infiltr

-0.68

POSITIVE LOGITS

Flickr

1.33

Getty

1.19

etc

1.07

ËĪ

1.05

AFP

1.05

sites

1.03

Gall

1.00

Kin

0.99

Staff

0.96

ãĤ¨ãĥ«

0.95

Activations Density 0.050%

financial figures or news headlines related to companies and individuals

occurrences of slashes or division symbols

No Comments

No Known Activations