INDEX

Explanations

references to violence and aggressive actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 10-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_10/width_16k/average_l0_77

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.10.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

icolon

-0.68

SerializedSize

-0.61

 Lázaro

-0.56

AndPassword

-0.55

Viitteet

-0.54

errHandler

-0.54

 RESTRICT

-0.53

jacent

-0.52

adjoint

-0.52

 télé

-0.52

POSITIVE LOGITS

 violent

0.86

 violence

0.86

 brawl

0.76

 injuring

0.75

 violento

0.73

 Violence

0.70

violence

0.68

Violence

0.68

 braw

0.64

Violent

0.63

Activations Density 0.405%

references to violence and aggressive actions

No Comments

No Known Activations