INDEX

Explanations

proper nouns related to individuals or entities

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

proper names and legal citations

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

 createState

-0.53

Slf

-0.53

Parcelize

-0.47

esModule

-0.44

Хьажоргаш

-0.43

/*

-0.42

nonatomic

-0.42

 consultato

-0.42

WriteLiteral

-0.41

]]=

-0.41

POSITIVE LOGITS

 betweenstory

0.64

 sessionFactory

0.60

征詢

0.51

 provecho

0.51

 Doubleday

0.51

 Configuración

0.51

 Stevenson

0.50

 McNamara

0.49

 befe

0.49

<=>

0.49

Activations Density 0.346%

proper nouns related to individuals or entities

proper names and legal citations

No Comments

No Known Activations