INDEX

Explanations

references to casualties and fatalities in violent events

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Configuration

Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post

Features

65,536

Data Type

float32

Hook Name

blocks.25.hook_resid_post

Hook Layer

Architecture

gated

Context Size

1,024

Dataset

Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 killers

-0.27

 killings

-0.24

kill

-0.23

 kills

-0.23

 killer

-0.22

Kill

-0.22

killer

-0.21

 kill

-0.21

 killed

-0.21

 killing

-0.20

POSITIVE LOGITS

 injured

0.32

 injury

0.32

 Injury

0.31

 wound

0.30

 wounded

0.30

 injuring

0.29

 injuries

0.29

 wounds

0.25

 seriously

0.24

inj

0.23

Activations Density 0.023%

references to casualties and fatalities in violent events

No Known Activations