INDEX

Explanations

hidden fields

This neuron detects references to hidden form inputs (e.g. HTML `<input type="hidden">` fields and related “hidden” form‐field mentions).

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

᾿

-1.12

FINISH

-0.96

EFIAPI

-0.93

 océano

-0.91

 systemctl

-0.89

графии

-0.88

 profeta

-0.87

词

-0.87

kindle

-0.87

 chré

-0.87

POSITIVE LOGITS

 hidden

1.24

 invis

1.05

hidden

1.00

 invisible

1.00

 important

0.96

all

0.94

key

0.93

 Invisible

0.91

 Hidden

0.89

赛事

0.81

Activations Density 0.007%