INDEX

Explanations

references to political norms and behaviors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

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

 ModelExpression

-0.63

Personendaten

-0.60

enterOuterAlt

-0.60

 purpoſe

-0.59

GEBURTSDATUM

-0.58

esModule

-0.57

 houſe

-0.56

 pleaſure

-0.55

 cauſe

-0.54

 disambiguazione

-0.53

POSITIVE LOGITS

 voluntarily

0.73

 voluntary

0.68

 Voluntary

0.56

 freiw

0.50

oluntary

0.43

 friv

0.41

 voluntarios

0.40

 vrijwilli

0.38

 willingly

0.37

 Volunt

0.36

Activations Density 0.147%

references to political norms and behaviors

No Comments

No Known Activations