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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

AndEndTag

-0.49

LookAnd

-0.48

TypedDataSet

-0.48

 almohada

-0.45

 telefónica

-0.45

GEBURTS

-0.44

dafx

-0.43

chufe

-0.43

 actionMode

-0.43

 المعيارى

-0.42

POSITIVE LOGITS

 Signalez

0.50

unknownFields

0.48

tangentMode

0.43

 Infórmanos

0.42

 EconPapers

0.42

ագրություններ

0.42

tanleria

0.39

 Италијани

0.37

 INTERESAR

0.36

 चीज़ों

0.35

Activations Density 0.000%

No Known Activations

This feature has no known activations.