INDEX

Explanations

references to evil or malevolent forces

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_1/width_16k/average_l0_102

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Paglinawan

-0.60

 monasterio

-0.58

 дописавши

-0.56

 STATS

-0.56

 Bumper

-0.56

FileDescriptor

-0.55

NewUrlParser

-0.54

 Pfar

-0.54

Демографія

-0.54

 pesado

-0.54

POSITIVE LOGITS

 Evil

1.54

EV

1.49

Evil

1.42

 evil

1.41

evil

1.36

EV

1.29

Ev

1.09

Ev

1.00

ev

0.90

ev

0.85

Activations Density 0.077%

references to evil or malevolent forces

No Comments

No Known Activations

references to evil or malevolent forces

No Comments

No Known Activations