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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 defensive

-0.36

copg

-0.35

łach

-0.34

 رز

-0.33

 boost

-0.32

طلحات

-0.32

fancy

-0.32

 socio

-0.32

疫

-0.32

 fancy

-0.31

POSITIVE LOGITS

modelBuilder

6.16

migrationBuilder

1.66

CreateModel

1.18

SequentialGroup

0.95

ScopeManager

0.93

ModelBuilder

0.92

 AssemblyTitle

0.92

DeleteBehavior

0.91

ViewFeatures

0.88

MigrationBuilder

0.86

Activations Density 0.000%

No Known Activations

This feature has no known activations.