INDEX

Explanations

terms related to geometric shapes, specifically circles and ellipses

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 1-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_1/width_16k/average_l0_105

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.1.hook_mlp_out

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

 circle

-2.72

 Circle

-2.38

circle

-2.38

Circle

-2.19

 circles

-2.14

 CIRCLE

-1.98

CIRCLE

-1.69

 Circles

-1.68

circles

-1.63

 círculo

-1.63

POSITIVE LOGITS

XmlAccessType

0.71

expandindo

0.68

 ReactDOM

0.62

verwijspagina

0.62

InsertCommand

0.61

 Amalgam

0.59

 Plung

0.57

rrggbb

0.56

:✨

0.56

endmodule

0.56

Activations Density 0.158%

terms related to geometric shapes, specifically circles and ellipses

No Comments

No Known Activations