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

ContentAlignment

-0.48

EndProject

-0.42

XmlEnum

-0.42

 ग्राहक

-0.41

ilead

-0.40

thâu

-0.39

 DialogResult

-0.38

>");

-0.38

uatu

-0.38

hells

-0.37

POSITIVE LOGITS

 ویکی‌پدیای

6.38

 ویکی‌پدیا

1.02

StoryboardSegue

0.81

 ویکی

0.79

hyrchwyd

0.67

存于互联网档案馆

0.66

 دیکھیے

0.63

 kasarigan

0.62

دانشنامهٔ

0.59

Архівовано

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.