INDEX

Explanations

No Explanations Found

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

DrawerToggle

-0.47

formik

-0.44

ArgsConstructor

-0.42

vů

-0.38

 HasFactory

-0.38

Cubit

-0.36

WaitGroup

-0.35

Демографія

-0.35

盡

-0.35

 yhte

-0.35

POSITIVE LOGITS

styleType

0.66

Portail

0.56

 gynhyrchwyd

0.52

RectangleBorder

0.50

 javafx

0.49

 BorderSide

0.49

 indoor

0.48

aser

0.48

 onDestroy

0.47

OPSIS

0.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.