INDEX

Explanations

even + adjective/adverb

This neuron activates on intensifying or qualifying adverbs and adverbial modifiers (e.g. “even,” “rather,” “outright,” “severe,” “in fact,” “completely”) that add emphasis or contrast.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

co

-1.35

 only

-1.32

-1.30

ng

-1.27

-1.26

me

-1.25

-1.23

nu

-1.20

POSITIVE LOGITS

 proveedores

1.36

 descrip

1.19

Stav

1.18

curios

1.16

 aplikace

1.14

 beetje

1.14

 LOGIC

1.13

cantara

1.13

 essais

1.13

 ママ

1.12

Activations Density 0.036%