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

 immagini

-0.85

Atsauces

-0.56

 images

-0.54

 nahilalakip

-0.54

 Images

-0.54

 imagens

-0.53

 mxArray

-0.52

Clik

-0.52

 imagery

-0.48

 Imágenes

-0.48

POSITIVE LOGITS

addLast

3.89

Expedia

0.93

 Expedia

0.67

$_['

0.46

AsUp

0.45

transQ

0.44

GTCX

0.42

первых

0.42

 làn

0.41

 khảo

0.40

Activations Density 0.000%

No Known Activations

This feature has no known activations.