INDEX

Explanations

numeric sequences and patterns typically associated with data formats or identifiers

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

 esternos

-0.46

WaitGroup

-0.45

 Infórmanos

-0.43

MediaStore

-0.42

السكان

-0.42

 Litchfield

-0.41

ThemeOverlay

-0.41

Demografía

-0.40

 jLabel

-0.40

tonsoft

-0.40

POSITIVE LOGITS

AnchorTagHelper

0.49

OGND

0.49

GOTREF

0.49

帖最后由

0.44

 onOptions

0.44

RefNanny

0.41

Tikang

0.39

 aDecoder

0.37

Coordonnées

0.37

 Biôgrafia

0.37

Activations Density 0.135%

numeric sequences and patterns typically associated with data formats or identifiers

No Comments

No Known Activations