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

AndEndTag

-0.94

RegressionTest

-0.86

WriteTagHelper

-0.83

parsedMessage

-0.80

ReusableCell

-0.77

RenderAtEndOf

-0.77

 snippetHide

-0.74

NameInMap

-0.73

[@BOS@]

-0.72

<unused3>

-0.72

POSITIVE LOGITS

enc

0.29

 Dienste

0.25

 enca

0.25

 secondaire

0.24

 Hens

0.23

 Services

0.23

clear

0.22

AnimationsModule

0.22

Thanks

0.22

 wzgl

0.22

Activations Density 0.000%

No Known Activations

This feature has no known activations.