INDEX

Explanations

describing actions or states

This neuron fires on morphologically inflected verbs—especially third-person singular present (“ameliorates”) and past participles (“driven,” “were”) in scientific contexts.

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

—

-1.55

).

-1.51

 those

-1.43

-1.38

ülerin

-1.37

-1.34

-1.33

].

-1.27

why

-1.26

-1.25

POSITIVE LOGITS

⇴

1.30

Slik

1.29

 världen

1.27

väg

1.24

を

1.24

 inspirasi

1.16

基本的に

1.15

 Different

1.14

 jajaran

1.14

 चीज़

1.13

Activations Density 0.435%