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

Personendaten

-0.71

 تضيفلها

-0.59

 onOptions

-0.53

AlterField

-0.52

ptonshire

-0.49

principalTable

-0.49

出版年

-0.48

 betweenstory

-0.47

duled

-0.46

 ſche

-0.44

POSITIVE LOGITS

astéroïdes

0.50

хьтан

0.49

Portale

0.42

 AssemblyCulture

0.39

ValueStyle

0.39

таратура

0.39

IsMutable

0.38

uxxxx

0.38

roaches

0.35

 femei

0.35

Activations Density 0.000%

No Known Activations

This feature has no known activations.