INDEX

Explanations

weighted avg

This neuron detects quantitative table headers and unit/metric labels (e.g. High, Avg, Change, Vol, Dist, energy, HRS).

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

 glä

-1.42

 sogenannten

-1.31

and

-1.29

información

-1.27

anız

-1.24

the

-1.20

Procedimento

-1.20

getNumberOf

-1.20

！！！！

-1.19

 برای

-1.17

POSITIVE LOGITS

yrs

1.88

 govt

1.63

 др

1.60

 approx

1.54

1.52

 např

1.46

 excluding

1.38

多かった

1.29

ってます

1.28

以外は

1.28

Activations Density 0.066%