INDEX

Explanations

terms and phrases related to evil and malevolent characteristics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 12-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

SceneManagement

-0.93

 noDo

-0.89

 дописавши

-0.86

InSection

-0.73

onView

-0.73

 GetEnumerator

-0.73

XtraReports

-0.73

 AspNetCore

-0.71

 Jockey

-0.70

NamedQueries

-0.70

POSITIVE LOGITS

 Evil

1.44

Evil

1.34

 evil

1.30

evil

1.16

 зло

0.79

 Smal

0.76

Mal

0.75

Bad

0.75

 evils

0.73

iastical

0.69

Activations Density 0.022%

terms and phrases related to evil and malevolent characteristics

No Comments

No Known Activations