INDEX

Explanations

names of people, sometimes with surrounding context like titles or initials

oai_token-act-pair · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 lume

-0.50

ueba

-0.50

 marche

-0.48

vedra

-0.48

 Soho

-0.47

 bross

-0.47

Pry

-0.47

zám

-0.46

barang

-0.46

kmale

-0.46

POSITIVE LOGITS

MemoryWarning

0.75

 يتيمه

0.72

 MainAxisSize

0.71

DoubleQuotes

0.68

niająca

0.66

0.63

Personendaten

0.63

Tembelea

0.62

HasForeignKey

0.61

AxisAlignment

0.60

Activations Density 0.591%

names of people, sometimes with surrounding context like titles or initials

No Comments

No Known Activations