INDEX

Explanations

references to evil and malevolent entities

oai_token-act-pair · gpt-4o-mini

New Auto-Interp

Configuration

chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.12.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

32,768

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Architecture

jumprelu

Context Size

1,024

Dataset

chanind/pile-uncopyrighted-gemma-1024-abbrv-1B

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 noDo

-0.88

 Ammon

-0.72

 дописавши

-0.71

 Upan

-0.69

webElementXpaths

-0.68

IsContent

-0.68

ChildScrollView

-0.66

 Bridges

-0.65

Bridges

-0.64

 تضيفلها

-0.62

POSITIVE LOGITS

 Evil

1.36

Evil

1.26

 evil

1.16

evil

1.13

 Devil

0.94

 Satan

0.89

 villain

0.87

Satan

0.85

 devil

0.83

villain

0.81

Activations Density 0.019%

references to evil and malevolent entities

No Comments

No Known Activations