INDEX

Explanations

or updated

The neuron fires on technical, uppercase format and protocol identifiers (e.g. “PCM,” “LPCM,” “embed,” “unsupported format,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

Also

1.30

Then

1.23

 then

1.19

 Also

1.13

งาน

1.09

然后

1.08

 także

1.07

 THEN

1.07

然後

1.06

then

1.05

POSITIVE LOGITS

 addirittura

1.72

 semplicemente

1.62

 equivalently

1.61

 simplemente

1.59

 simply

1.54

 simplement

1.47

 alternatively

1.46

 simplesmente

1.45

 просто

1.42

 bahkan

1.36

Activations Density 0.563%