INDEX

Explanations

email addresses or related phrases

oai_token-act-pair · gpt-3.5-turbo

words or phrases related to contact information and email protection

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 elig

-0.35

stown

-0.33

alse

-0.32

ington

-0.32

oult

-0.31

Js

-0.30

indal

-0.30

 Colon

-0.30

irie

-0.29

 Plaintiff

-0.29

POSITIVE LOGITS

ccording

0.35

ãĥ´ãĤ¡

0.33

 cheesy

0.33

ĳå£«

0.32

 frustrating

0.32

 sclerosis

0.31

quickShipAvailable

0.31

 impetus

0.30

?????-?????-

0.30

 heartbreaking

0.30

Activations Density 1.393%

email addresses or related phrases

words or phrases related to contact information and email protection

No Comments

No Known Activations