INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-res/layer_12/width_1m/average_l0_107

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

1,048,576

Data Type

float32

Hook Name

blocks.12.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

 udaler

-0.61

ništvo

-0.57

Gambas

-0.56

 Infórmanos

-0.56

τή

-0.56

consulté

-0.55

Бахар

-0.55

MessageOf

-0.55

titutions

-0.54

Datuak

-0.53

POSITIVE LOGITS

://"

0.59

 Numerade

0.54

bewerken

0.54

DebuggerNonUser

0.52

SBATCH

0.51

UnitTesting

0.49

tagext

0.48

anggil

0.48

ControllerAdvice

0.48

åne

0.48

Activations Density 0.000%

No Known Activations

This feature has no known activations.