INDEX

Explanations

references to people involved in leadership or public service roles

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

former titles and names

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

ngdoc

-0.41

contentLoaded

-0.35

yakarta

-0.34

 aktuell

-0.32

badah

-0.32

ckså

-0.32

quième

-0.32

hObject

-0.31

这就

-0.30

édr

-0.30

POSITIVE LOGITS

Former

0.69

ScopeManager

0.67

 Former

0.66

former

0.65

 former

0.63

Retired

0.56

 retirees

0.53

dropna

0.52

retired

0.52

 ehemaligen

0.52

Activations Density 0.021%

references to people involved in leadership or public service roles

former titles and names

No Comments

No Known Activations