INDEX

Explanations

values or identifiers related to coding or data structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

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

AndEndTag

-0.66

DropTable

-0.64

MLLoader

-0.54

 voeten

-0.53

⇑

-0.50

ایج

-0.50

 ModelExpression

-0.49

:✨

-0.49

tover

-0.49

andar

-0.49

POSITIVE LOGITS

SBATCH

0.64

Atsauces

0.59

stateProvider

0.57

metheus

0.53

 CommonModule

0.52

DriverManager

0.52

 useCallback

0.50

 NSObject

0.50

IFS

0.47

tonsoft

0.47

Activations Density 0.114%

values or identifiers related to coding or data structures

No Comments

No Known Activations