INDEX

Explanations

auxiliary verbs followed by verbs

The neuron detects disclaimer or warning statements (e.g. “Any bugs are your problem,” “modifications … will be lost,” “must retain …”) in code or documentation.

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

このまま

-1.55

 guarn

-1.46

 discrepan

-1.45

 desg

-1.42

 exista

-1.39

 algu

-1.33

 captur

-1.31

 CASO

-1.31

 notor

-1.31

によっては

-1.30

POSITIVE LOGITS

 also

1.72

 with

1.66

 would

1.48

 raczej

1.48

 that

1.43

 including

1.42

 encuentros

1.41

 včetně

1.35

этому

1.31

jį

1.30

Activations Density 0.073%