INDEX

Explanations

PCIe Gen 4

The neuron strongly activates on numeric tokens (e.g. version numbers, years, measurements), i.e. any token containing digits.

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

シャー

0.56

 Toward

0.55

HSE

0.55

 residence

0.55

પતિ

0.55

 Cass

0.54

 Irving

0.54

 Valeria

0.54

 Trunk

0.54

 Haynes

0.52

POSITIVE LOGITS

nodes

0.59

pidos

0.58

ud

0.54

 núcleos

0.54

 participantes

0.53

 könnten

0.53

inl

0.53

 телефо

0.52

={()

0.52

 सुरू

0.51

Activations Density 0.000%