INDEX

Explanations

code keywords and symbols

The neuron flags calls to visibility‐toggling methods (like hide(), show(), slideUp(), etc.) in code.

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

 IconData

-1.08

和

-0.85

 slutet

-0.85

 päivä

-0.85

 språk

-0.84

 commemoration

-0.84

 petra

-0.84

ambiguation

-0.84

 január

-0.83

ustre

-0.83

POSITIVE LOGITS

 pousse

0.94

辍

0.94

 aktiv

0.94

nor

0.91

 febru

0.88

aktiv

0.85

 trasm

0.84

mieszkanie

0.82

 norvég

0.81

 cesta

0.80

Activations Density 0.001%