INDEX

Explanations

phrases indicating a suggestion or consideration, usually starting with "Keep in mind"

oai_token-act-pair · gpt-3.5-turbo

phrases that emphasize the need to consider or remember specific information

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

MpServer

-0.68

inals

-0.67

iel

-0.64

QL

-0.63

git

-0.62

EStreamFrame

-0.62

blogspot

-0.62

 ribbon

-0.61

cgi

-0.60

rib

-0.60

POSITIVE LOGITS

 lest

0.78

anwhile

0.69

pad

0.67

 Wond

0.64

erning

0.63

fulness

0.63

how

0.62

othal

0.62

swer

0.62

 caveat

0.62

Activations Density 0.019%

phrases indicating a suggestion or consideration, usually starting with "Keep in mind"

phrases that emphasize the need to consider or remember specific information

No Comments

No Known Activations