INDEX

Explanations

references to notable or exemplary individuals, cultural icons, or popular entities

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

best of the best

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

intios

-0.48

 snippetHide

-0.47

Diweddarwch

-0.43

ंदीखरीदारी

-0.40

 disponibilités

-0.39

 Muslim

-0.38

 pylint

-0.37

setCellStyle

-0.37

 numberOfRows

-0.37

OptionsMenu

-0.35

POSITIVE LOGITS

 kaarangay

0.57

 pungkas

0.57

 pungkasnya

0.56

 gloire

0.55

ükemmel

0.54

SourceChecksum

0.54

 eenig

0.54

 Schönheit

0.53

 paixão

0.53

 extraordinaire

0.52

Activations Density 0.022%

references to notable or exemplary individuals, cultural icons, or popular entities

best of the best

No Comments

No Known Activations