INDEX

Explanations

references to specific individuals and notable nouns

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 AssemblyCulture

-0.50



-0.50

 Exactos

-0.50

ReusableCell

-0.47

Ӕ

-0.43

routeProvider

-0.43

таратура

-0.42

FetchType

-0.42

didSet

-0.42

 Chwiliwch

-0.41

POSITIVE LOGITS

 betweenstory

0.46

舺

0.41

хьтан

0.40

 disponibilités

0.39

Lone

0.39

 Normdatei

0.38

 Transfermarkt

0.38

])));

0.37

edal

0.37

 Meksiku

0.37

Activations Density 0.046%

references to specific individuals and notable nouns

No Comments

No Known Activations