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

webElementXpaths

-0.63

SourceChecksum

-0.62

BufferException

-0.50

 mags

-0.46

="@+

-0.45

ксу

-0.45

RepeatedField

-0.44

[]>(

-0.43

 Италијани

-0.43

enterOuterAlt

-0.41

POSITIVE LOGITS

ecore

4.66

 kasarigan

0.54

+#+

0.49

portál

0.49

 NSCoder

0.48

OGND

0.46

ROIT

0.45

Bissau

0.44

 BoxDecoration

0.43

 endblock

0.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.