INDEX

Explanations

No Explanations Found

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 unspeak

-1.68

 belliger

-1.63

 disgra

-1.52

 exasper

-1.48

 horrend

-1.45

 ruinous

-1.44

 odious

-1.42

 scound

-1.39

 despotism

-1.39

 ineffectual

-1.37

POSITIVE LOGITS

<bos>

12.17

expandindo

2.02

GEBURTSDATUM

2.02

 betweenstory

1.90

Autoritní

1.89

تقاوى

1.87

 Administrativna

1.73

 Италијани

1.68

 '\\;'

1.67

 autorytatywna

1.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.