INDEX

Explanations

references to land, particularly in the context of geographical or scientific discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.17.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

kloped

-0.67

endphp

-0.66

Искәрмәләр

-0.65

ftagPool

-0.57

Демографія

-0.56

Gothic

-0.52

存于互联网档案馆

-0.51

قایناق‌لار

-0.51

 UIKit

-0.51

Artificial

-0.51

POSITIVE LOGITS

 Land

0.94

 LAND

0.89

Land

0.88

 land

0.76

 Lande

0.67

Lan

0.66

LAN

0.63

LAND

0.61

land

0.60

Lan

0.60

Activations Density 0.027%

references to land, particularly in the context of geographical or scientific discussions

No Comments

No Known Activations