INDEX

Explanations

phrases related to violent acts resulting in death

oai_token-act-pair · gpt-3.5-turbo

references to violent deaths and murder

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

æ©Ł

-0.81

inventoryQuantity

-0.75

ĨĴ

-0.73

rium

-0.71

Factor

-0.71

soType

-0.69

iago

-0.66

annis

-0.66

omaly

-0.66

 Seller

-0.65

POSITIVE LOGITS

 unarmed

0.96

 murdering

0.92

 murdered

0.81

 senseless

0.80

 rampage

0.78

 murder

0.78

 indiscrim

0.76

 gunned

0.76

 spree

0.75

 revenge

0.75

Activations Density 0.135%

phrases related to violent acts resulting in death

references to violent deaths and murder

No Comments

No Known Activations