INDEX

Explanations

proper nouns related to people, organizations, or specific terms

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

names and places that are entities

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

Aiheesta

-0.44

 NSCoder

-0.40

prnewswire

-0.40

UpInside

-0.40

HtmlAttribute

-0.40

spo

-0.35

日閲覧

-0.34

medriver

-0.33

setupUi

-0.32

spo

-0.32

POSITIVE LOGITS

 betweenstory

0.54

 Meksiku

0.51

fillType

0.50

AsUp

0.49

msgTypes

0.47

ThemeOverlay

0.47

 Grüsse

0.47

 Chwiliwch

0.46

ʺ

0.45

webElementGuid

0.45

Activations Density 0.042%

proper nouns related to people, organizations, or specific terms

names and places that are entities

No Comments

No Known Activations