INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

Vim

-0.48

osocial

-0.47

SBATCH

-0.45

 ProtoMessage

-0.44

vim

-0.43

 препратки

-0.43

toHaveBeen

-0.43

extAlignment

-0.42

kum

-0.41

orsese

-0.41

POSITIVE LOGITS

WriteAttribute

0.67

fjspx

0.54

saraba

0.48

 nahilalakip

0.48

AutoSize

0.46

ftagPool

0.44

richTextPanel

0.44

setVerticalGroup

0.43

 fotografico

0.42

 CreateTagHelper

0.42

Activations Density 0.000%

No Known Activations

This feature has no known activations.