INDEX

Explanations

terms related to descriptions and characterizations of concepts or objects

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

帖最后由

-0.45

fromnode

-0.38

featureID

-0.38

嬸

-0.34

Auszeichnungen

-0.34

 Gunakan

-0.33

GeoNames

-0.32

Libs

-0.32

Viited

-0.32

tagHelperRunner

-0.31

POSITIVE LOGITS

how

0.68

 representing

0.57

 represents

0.56

!*\

0.54

 means

0.54

 означает

0.54

 denotes

0.53

 signifies

0.52

 signify

0.52

 represent

0.52

Activations Density 0.119%

terms related to descriptions and characterizations of concepts or objects

No Comments

No Known Activations