INDEX

Explanations

structured content and elements that indicate organization or formatting in text

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

titles and headings followed by common words

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

 يتيمه

-0.60

UrlResolution

-0.50

SourceChecksum

-0.50

 Chwiliwch

-0.50

 Administrativna

-0.48

MigrationBuilder

-0.47

 AssemblyCulture

-0.47

 typelib

-0.47

WebElementEntity

-0.46

يكب

-0.46

POSITIVE LOGITS

bear

0.52

 Emer

0.44

 Ride

0.43

AutoScale

0.43

✭✭

0.42

 Blog

0.41

Zo

0.41

ride

0.41

 Menu

0.40

Activations Density 0.115%

structured content and elements that indicate organization or formatting in text

titles and headings followed by common words

No Comments

No Known Activations