INDEX

Explanations

phrases or statements directed towards someone

oai_token-act-pair · gpt-3.5-turbo

words related to personal pronouns and references to people

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 vulner

-0.63

emale

-0.61

],[

-0.53

 Azerb

-0.53

edIn

-0.52

 Afgh

-0.52

 surn

-0.50

 horizont

-0.50

 Borders

-0.50

ilaterally

-0.49

POSITIVE LOGITS

amins

0.65

ICE

0.63

Fish

0.60

STAR

0.59

 Entered

0.59

quickShipAvailable

0.59

 ROCK

0.58

cop

0.58

CRIPTION

0.58

weed

0.57

Activations Density 0.313%

phrases or statements directed towards someone

words related to personal pronouns and references to people

No Comments

No Known Activations