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

 localizar

-0.54

 Vogt

-0.51

 Metzger

-0.50

webElementXpaths

-0.50

 Administrativna

-0.50

AlterField

-0.44

 Furcht

-0.44

 interposed

-0.44

 Rptr

-0.43

 jantung

-0.43

POSITIVE LOGITS

ArgumentParser

0.44

 colleges

0.43

 Colleges

0.43

Наводи

0.43

@"/

0.42

structors

0.42

setVerticalGroup

0.42

 pleaſure

0.41

UseVisualStyle

0.40

 universities

0.40

Activations Density 0.000%

No Known Activations

This feature has no known activations.