INDEX

Explanations

the word "to" and words and short phrases that often come before it.

oai_token-act-pair · gemini-2.0-flash

to

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_20/width_16k/average_l0_11

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

 esternos

-0.76

AndEndTag

-0.67

 estekak

-0.67

enumi

-0.66

GeneratedCode

-0.64

HasForeignKey

-0.63

enumii

-0.57

achau

-0.54

 reú

-0.54

 nawr

-0.53

POSITIVE LOGITS

openha

0.55

 Gazetteer

0.53

">#

0.51

 artesan

0.50

#}

0.49

 EClass

0.48

 désolés

0.48

 Candel

0.47

felizmente

0.47

Blurb

0.47

Activations Density 0.451%

the word "to" and words and short phrases that often come before it.

to

No Comments

No Known Activations