INDEX

Explanations

phrases related to causing harm or death

oai_token-act-pair · gpt-3.5-turbo

references to killing or violent actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

BuyableInstoreAndOnline

-0.76

Cola

-0.76

oday

-0.73

ĸļ

-0.73

ORY

-0.73

 Failure

-0.67

DragonMagazine

-0.65

 Municip

-0.65

fty

-0.65

wcsstore

-0.63

POSITIVE LOGITS

joy

1.03

 spree

0.92

switch

0.86

mails

0.85

off

0.81

 civilians

0.80

ously

0.76

 whales

0.75

mong

0.75

lords

0.75

Activations Density 0.059%

phrases related to causing harm or death

references to killing or violent actions

No Comments

No Known Activations