INDEX

Explanations

dispute and reconciliation

The neuron fires on terms that indicate disagreement, conflict, or being at odds (e.g. “rift,” “odds,” “disagreement,” “ostracized”).

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

tisp

-1.38

 рит

-1.37

ussure

-1.31

すら

-1.28

 inspect

-1.27

 afficher

-1.26

 поход

-1.26

 serons

-1.26

фессиона

-1.25

ştır

-1.24

POSITIVE LOGITS

 tuvimos

1.39

quel

1.34

 discussing

1.30

erweise

1.20

 تو

1.20

=()=>{

1.19

1.17

 tuttavia

1.16

scape

1.16

Activations Density 0.025%