INDEX

Explanations

's and 't contractions

The neuron detects numerical and quantitative tokens (digits, percentages, and similar numeric expressions).

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

 kuchni

-1.45

 oświet

-1.40

 mojej

-1.36

 zunächst

-1.35

 trusty

-1.34

 obligé

-1.34

 tył

-1.33

 przód

-1.33

 moim

-1.33

bbbb

-1.31

POSITIVE LOGITS

 only

2.09

can

1.53

 will

1.52

1.49

now

1.47

in

1.42

 there

1.38

//!

1.37

}{

1.32

 needed

1.31

Activations Density 0.060%