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

 Administrativna

-0.62

GEBURTSDATUM

-0.58

$_['

-0.56

GEBURTS

-0.56

 ویکی‌پدی

-0.54

SharedDtor

-0.52

DotNetBar

-0.50

MigrationBuilder

-0.48

jsxFileName

-0.47

libft

-0.47

POSITIVE LOGITS

 archive

0.71

 Archive

0.66

archive

0.64

Archive

0.60

 archives

0.57

 Archives

0.50

archives

0.49

 archivio

0.48

 arkiv

0.47

 archiving

0.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.