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

 ModelExpression

-0.56

Hentet

-0.54

 للاسماء

-0.47

GEBURTS

-0.46

InjectAttribute

-0.46

 pinulongan

-0.44

 defaultstate

-0.44

MessageOf

-0.43

protoimpl

-0.43

MeasureSpec

-0.43

POSITIVE LOGITS

leski

1.40

sidemargin

0.67

AccessorTable

0.54

/**

0.45

orka

0.44

CreateIndex

0.44

apist

0.43

locales

0.43

 BytesLike

0.42

 Loki

0.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.