INDEX

Explanations

even including across languages

The neuron detects numeric quantity expressions (numbers often paired with units, like durations, distances, counts, etc.).

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

but

-1.14

or

-1.06

 would

-1.03

ūros

-1.02

for

-1.02

 could

-0.96

 becomes

-0.95

 that

-0.95

 only

-0.93

createClass

-0.90

POSITIVE LOGITS

Including

1.38

without

1.36

 sogar

1.34

 навіть

1.30

 даже

1.24

 quelquefois

1.22

 problemlos

1.20

 bahkan

1.20

 jopa

1.19

 incluso

1.16

Activations Density 0.061%