INDEX

Explanations

references to TV shows and news programs

oai_token-act-pair · gemini-2.0-flash

TV shows

np_max-act · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_8/width_16k/average_l0_52

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.8.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tagext

-0.49

agu

-0.48

 truncate

-0.47

λου

-0.47

 Hald

-0.46

..\..\

-0.45

بوابة

-0.44

úgó

-0.44

amazonaws

-0.44

 telegraph

-0.44

POSITIVE LOGITS

<bos>

0.61

 transfieras

0.61

 betweenstory

0.59

IndentedString

0.54

tagHelper

0.53

 Italijani

0.53

 Paglinawan

0.52

 transcur

0.51

irchen

0.50

elemField

0.49

Activations Density 0.611%

references to TV shows and news programs

TV shows

No Comments

No Known Activations