INDEX

Explanations

references to user interface elements and structure, particularly related to menus and selections

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

menu or table headers

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

 esponja

-0.46

 felices

-0.45

بوابة

-0.39

 nôtre

-0.38

 libremente

-0.38

 parezca

-0.38

 Debido

-0.38

 inmediato

-0.38

endaft

-0.37

 idéia

-0.37

POSITIVE LOGITS

 queſta

0.58

0.53

SourceChecksum

0.52

 otomatig

0.49

 Meksiku

0.48

 Италијани

0.48

ICEF

0.48

 Мексичка

0.47

 Babel

0.46

 kohdetta

0.46

Activations Density 0.124%

references to user interface elements and structure, particularly related to menus and selections

menu or table headers

No Comments

No Known Activations