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

⤹

-0.43

+#+#

-0.42

ắn

-0.42

RuleContext

-0.41

estacks

-0.40

aro

-0.39

 становника

-0.39

COUVER

-0.38

gulp

-0.38

HtmlAttribute

-0.37

POSITIVE LOGITS

HasAnnotation

1.70

 kasarigan

0.68

 Roskov

0.59

migrationBuilder

0.54

 ligiloj

0.52

 nahilalakip

0.46

 poffible

0.45

 Infórmanos

0.42

Datuak

0.42

 незавершена

0.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.