INDEX

Explanations

headings or structured elements in text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

хьтан

-0.43

delwed

-0.42

 EXPEND

-0.40

 CULTURAL

-0.39

 autorytatywna

-0.39

Diweddarwch

-0.39

 STANDARDS

-0.39

audiovisuel

-0.38

 autoridad

-0.38

 TRAB

-0.36

POSITIVE LOGITS

 ProtoMessage

0.56

 defaultstate

0.51

)++;

0.48

Tembelea

0.47

 nakalista

0.46

 HasFactory

0.46

 joaat

0.46

########.

0.43

énario

0.43

GBK

0.43

Activations Density 0.013%

headings or structured elements in text

No Comments

No Known Activations