INDEX

Explanations

expressing uncertainty or a thought

This neuron detects informal discourse markers and hedging phrases (e.g. “you know,” “I think,” “one might wonder,” 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

 also

-1.45

its

-1.38

 також

-1.36

 however

-1.12

 then

-1.12

 église

-1.10

<bos>

-1.08

그리고

-1.07

atized

-1.05

erdings

-1.04

POSITIVE LOGITS

 okreś

1.11

 précédentes

1.05

 mögen

1.05

ことはない

1.04

已经

1.04

ново

1.03

 tässä

1.02

 funktionieren

0.98

 mittlerweile

0.97

ಒ

0.97

Activations Density 0.111%