INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_18/width_65k/average_l0_88

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.18.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

LayoutStyle

-0.77

HtmlAttribute

-0.72

tamol

-0.66

 Administrativna

-0.63

awtextra

-0.62

mandu

-0.59

mtable

-0.58

دانشنامهٔ

-0.57

Hauptartikel

-0.57

Hentet

-0.57

POSITIVE LOGITS

ecma

0.49

ยาก

0.47

OfYear

0.44

âm

0.44

ye

0.43

InteropServices

0.43

Sklici

0.42

 immediate

0.42

kháu

0.41

 basket

0.41

Activations Density 0.000%

No Known Activations

This feature has no known activations.