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

transQ

-0.49

 FontWeight

-0.46

VertexAttrib

-0.46

انتهای

-0.42

 onAnimation

-0.40

yarnpkg

-0.39

UIControlState

-0.38

providedIn

-0.38

 createSlice

-0.38

 BigNumberish

-0.38

POSITIVE LOGITS

rtype

2.53

QMetaType

0.53

type

0.52

httphttps

0.51

addContainerGap

0.49

 typelib

0.48

returnType

0.47

+#+

0.47

 type

0.47

dataType

0.46

Activations Density 0.000%

No Known Activations

This feature has no known activations.