INDEX

Explanations

calculating probabilities with replacement

The neuron is specialized to spot standalone numeric tokens (strings of digits).

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

 where

-1.13

رفة

-1.13

latimes

-1.08

Flere

-1.05

donde

-1.03

Suivez

-1.03

YMS

-1.01

ではまた

-1.00

')));

-0.99

Donde

-0.98

POSITIVE LOGITS

1.23

1.03

1.02

1.01

0.98

0.97

entr

0.88

像是

0.88

0.86

Activations Density 0.013%