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

-1.82

 horrend

-1.61

 ineffec

-1.61

 ruinous

-1.59

 belliger

-1.59

 laug

-1.50

 unlaw

-1.50

 shenan

-1.49

 unavoid

-1.48

 miscon

-1.48

POSITIVE LOGITS

<bos>

11.00

Autoritní

2.27

GEBURTSDATUM

2.24

expandindo

2.05

 autorytatywna

1.94

 betweenstory

1.93

 '\\;'

1.87

 Administrativna

1.80

تقاوى

1.79

LookAnd

1.77

Activations Density 0.000%

No Known Activations

This feature has no known activations.