INDEX

Explanations

terms related to visual processing and rendering in application code

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

 final

-0.37

額

-0.33

 asesinado

-0.31

 döv

-0.31

tablir

-0.30

zeiger

-0.29

 genomen

-0.28

 desaparecido

-0.28

 raro

-0.28

 llegado

-0.28

POSITIVE LOGITS

 ModelRenderer

0.89

parsedMessage

0.80

featureID

0.75

 surla

0.73

setVerticalGroup

0.73

BufferException

0.71

IsMutable

0.70

WithIOException

0.69

RegressionTest

0.69

+#+#

0.68

Activations Density 0.007%

terms related to visual processing and rendering in application code

No Comments

No Known Activations