INDEX

Explanations

exceptions, error, invalid, returns

The neuron detects mentions of error or exception identifiers (names of errors/exceptions).

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

┍

-1.38

ườn

-1.22

 функция

-1.21

 никогда

-1.21

 aldri

-1.17

 Pfal

-1.17

钥匙

-1.16

Sklici

-1.16

 Aufen

-1.15

✴

-1.15

POSITIVE LOGITS

1.63

we

1.54

We

1.41

1.23

 terzo

1.21

\_

1.20

if

1.19

 комнату

1.16

or

1.13

 puisqu

1.13

Activations Density 0.005%