INDEX

Explanations

segments related to introductions or structured content in documents

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

introduction or book or text

np_acts-logits-general · gemini-2.5-flash-lite

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 betweenstory

-0.52

transQ

-0.51

Personensuche

-0.46

*/),

-0.45

.*")]

-0.45

󠁴

-0.44

autaire

-0.43

Datuak

-0.42

RectangleBorder

-0.42

 deleteById

-0.42

POSITIVE LOGITS

Personendaten

0.56

ंदीखरीदारी

0.51

hyrchwyd

0.49

niająca

0.48

Jeografia

0.47

 above

0.46

ViewImports

0.45

новништво

0.45

таратура

0.45

FunctionFlags

0.42

Activations Density 0.013%

segments related to introductions or structured content in documents

introduction or book or text

No Comments

No Known Activations