INDEX

Explanations

contextual markers or phrases indicating the beginning of a new section or topic

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

viamo

-0.60

OLIC

-0.59

érc

-0.57

álva

-0.57

ervis

-0.55

íqu

-0.53

ģ

-0.53

}}_{

-0.53

IFIED

-0.52

sprites

-0.52

POSITIVE LOGITS

 estekak

0.90

jsxFileName

0.73

Havolalar

0.68

μβρίου

0.65

 lenker

0.64

canestro

0.63

UnsafeEnabled

0.63

Personensuche

0.63

 ivelany

0.61

évaluateur

0.60

Activations Density 0.057%

contextual markers or phrases indicating the beginning of a new section or topic

No Comments

No Known Activations