INDEX

Explanations

logging and configuration

This neuron fires on mentions of software module or package names (e.g. code library identifiers).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 tolerated

1.03

 tolerate

0.95

 Toler

0.95

toler

0.92

 toler

0.92

 tolerance

0.90

 Tolerance

0.88

 tolerable

0.83

tolerance

0.81

Tolerance

0.80

POSITIVE LOGITS

 save

0.60

 Save

0.58

save

0.57

Evo

0.55

 minions

0.53

 trusted

0.52

 উৎপাদনের

0.52

 SAVE

0.51

 store

0.51

 clos

0.50

Activations Density 0.023%