INDEX

Explanations

phrases related to legal cases or investigations

oai_token-act-pair · gpt-3.5-turbo

references to allegations of criminal behavior, particularly involving sexual offenses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Tokens

-0.88

DragonMagazine

-0.86

terms

-0.84

izons

-0.82

prints

-0.81

eday

-0.80

hest

-0.78

hers

-0.76

ername

-0.74

SpaceEngineers

-0.74

POSITIVE LOGITS

 abuse

1.28

 misconduct

1.24

 vandalism

1.23

 harassment

1.19

 discrimination

1.18

 violence

1.15

 wrongful

1.14

 criminality

1.12

 wrongdoing

1.12

 rape

1.11

Activations Density 0.186%

phrases related to legal cases or investigations

references to allegations of criminal behavior, particularly involving sexual offenses

No Comments

No Known Activations