INDEX

Explanations

names of individuals and their significance in discussions or statements

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

names followed by attribution

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

featureID

-0.93

GEBURTSDATUM

-0.86

oa̍t

-0.84

parsedMessage

-0.81

endphp

-0.77

 estekak

-0.75

хьтан

-0.72

 ویکی‌پدی

-0.71

NameInMap

-0.70

 Италијани

-0.70

POSITIVE LOGITS

 vaiz

0.33

 Trabaj

0.32

ambung

0.31

 Sosten

0.31

wyg

0.31

 Wolken

0.30

 Scheiben

0.30

 Welt

0.29

 geht

0.29

Dữ

0.29

Activations Density 0.053%

names of individuals and their significance in discussions or statements

names followed by attribution

No Comments

No Known Activations