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

 للاسماء

-1.10

 doInBackground

-1.05

tagHelperRunner

-0.71

SPATH

-0.71

Personensuche

-0.56

 kasarigan

-0.51

transQ

-0.50

Diweddarwch

-0.48

 GridLayout

-0.47

principalTable

-0.46

POSITIVE LOGITS

destroyAll

2.23

PageInfo

1.48

 WaitForSeconds

1.13

WaitGroup

0.85

PageContext

0.65

pageInfo

0.52

ByVersion

0.52

sizeCache

0.49

dafx

0.47

getDoctrine

0.46

Activations Density 0.000%

No Known Activations

This feature has no known activations.