INDEX

Explanations

phrases related to customization or personalized features

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 12-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_12/width_16k/average_l0_130

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.12.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

Kjelder

-0.62

 viewDidLoad

-0.47

SqlDataReader

-0.47

Autoritní

-0.46

 phoneNumber

-0.45

XmlAccessType

-0.43

 pinulongan

-0.42

migrationBuilder

-0.42

 wzgl

-0.42

 ligiloj

-0.41

POSITIVE LOGITS

 custom

1.65

 Custom

1.60

Custom

1.49

custom

1.45

 CUSTOM

1.38

 customized

1.37

CUSTOM

1.32

setCustom

1.31

 Customized

1.27

 customize

1.26

Activations Density 0.102%

phrases related to customization or personalized features

No Comments

No Known Activations