INDEX

Explanations

indications of perceived manipulation or mistreatment

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

出版年

-0.51

 nahilalakip

-0.48

 AssemblyTitle

-0.45

//};

-0.41

eqref

-0.39

 vides

-0.38

 ویکی‌پدیا

-0.38

 conex

-0.38

 muối

-0.37

 vixion

-0.35

POSITIVE LOGITS

лтемелер

0.46

MemoryWarning

0.46

CppMethod

0.45

 told

0.45

enterOuterAlt

0.44

IsContent

0.44

fromnode

0.43

 indoctr

0.42

pense

0.41

ridgeshire

0.41

Activations Density 0.100%

indications of perceived manipulation or mistreatment

No Comments

No Known Activations