INDEX

Explanations

references to North America

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.hook_mlp_out

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

 America

-0.90

America

-0.75

 AMERICA

-0.67

 america

-0.63

 "..\..\

-0.60

 aapt

-0.52

america

-0.52

WebServlet

-0.52

toHaveBeen

-0.51

 OnInit

-0.50

POSITIVE LOGITS

цездатний

0.68

 meanwhile

0.63

 cappuccio

0.62

 insegna

0.60

Опис

0.60

ensed

0.60

Ruh

0.60

/:

0.59

Presse

0.59

Activations Density 0.013%

references to North America

No Comments

No Known Activations