INDEX

Explanations

conflict, contradiction, and precedence

overriding and superseding

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 researches

0.46

 initial

0.42

 inicial

0.42

 Hodge

0.40

 sterling

0.40

 inital

0.39

 requisite

0.39

Alh

0.38

ris

0.38

 solvable

0.38

POSITIVE LOGITS

contrad

0.56

 contradictory

0.56

 inconsistency

0.56

矛盾

0.55

 conflicting

0.55

 contradiction

0.55

Conflicts

0.55

Conflict

0.51

 priorité

0.51

override

0.50

Activations Density 0.057%