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

Diweddarwch

-0.54

bitField

-0.47

AntiForgeryToken

-0.39

sizeCache

-0.38

iddish

-0.34

 chart

-0.33

 butcher

-0.33

="@+

-0.32

tonsoft

-0.32

 however

-0.31

POSITIVE LOGITS

 pinulongan

0.60

Lähteet

0.59

Kaynakça

0.57

HtmlAttribute

0.52

发表于

0.50

 autorytatywna

0.50

 AssemblyTitle

0.49

 محفوظة

0.47

webElementXpaths

0.47

providedIn

0.47

Activations Density 0.000%

No Known Activations

This feature has no known activations.