INDEX

Explanations

conditional questions

The neuron detects first‐person expressions of desire or intention (e.g. “I want to …,” “asked if I was …”).

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

 dagens

-1.38

 today

-1.37

 Autó

-1.20

 aujourd

-1.13

 yesterday

-1.11

ॉग

-1.06

 نداره

-1.05

AILY

-1.04

今日も

-1.01

 Giappone

-1.01

POSITIVE LOGITS

if

1.27

 если

1.23

 would

1.16

 voidaan

1.13

 jeśli

1.13

 olisi

1.13

 якщо

1.13

 eğer

1.10

年には

1.05

之事

1.03

Activations Density 0.023%