INDEX

Explanations

confidentiality, employment, contracts, competition law

The neuron detects mentions of confidential or proprietary information—especially terms like “trade secrets,” “confidential information,” and related legal secrecy language.

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

 batterie

-1.44

<td>

-1.25

 alimente

-1.23

vei

-1.23

 hiver

-1.23

 vider

-1.22

 apparence

-1.20

慙

-1.20

 Mockito

-1.19

:''

-1.16

POSITIVE LOGITS

 ALWAYS

1.34

If

1.27

as

1.25

 solely

1.23

 Because

1.21

 their

1.21

 Even

1.20

at

1.20

 purposefully

1.19

 that

1.17

Activations Density 0.045%