INDEX

Explanations

references to death and murder-related events

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

WithIOException

-0.40

]")]

-0.39

 ActionListener

-0.39

namentales

-0.38

 kasarigan

-0.35

 AttributeSet

-0.34

ДД

-0.34

Спасылкі

-0.34

prepareStatement

-0.34

kkue

-0.33

POSITIVE LOGITS

 death

0.65

 deaths

0.64

 mourut

0.57

 asesinado

0.56

 died

0.54

death

0.54

 deceased

0.54

 tragic

0.52

 victims

0.52

 surla

0.51

Activations Density 0.091%

references to death and murder-related events

No Comments

No Known Activations