INDEX

Explanations

references to terrorist attacks and related incidents

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 36-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_36/width_16k/average_l0_120

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.36.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

createCell

-0.32

AppMethodBeat

-0.31

IllegalArgument

-0.30

 messing

-0.29

 autorytatywna

-0.29

RefNanny

-0.28

 irré

-0.27

 insect

-0.27

bed

-0.26

 aventure

-0.26

POSITIVE LOGITS

 deadliest

0.91

 massacre

0.90

 shootings

0.88

 gunman

0.86

 tragedy

0.84

 Massacre

0.80

 massacres

0.79

CodeAttribute

0.78

 shooter

0.75

 tragedies

0.75

Activations Density 0.269%

references to terrorist attacks and related incidents

No Comments

No Known Activations