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

ments

-0.37

 verbund

-0.36

setVerticalGroup

-0.35

 själva

-0.35

inken

-0.33

 أد

-0.33

 eingestellt

-0.33

-0.32

POSITIVE LOGITS

 Bailly

1.94

 незавершена

0.78

 AssemblyCulture

0.73

ValueStyle

0.68

AndEndTag

0.63

 Signalez

0.63

 Taktlose

0.63

 Exactos

0.62

 disambiguazione

0.62

Gambas

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.