INDEX

Explanations

Next.js

The neuron strongly activates on action‐oriented verbs that introduce steps or transitions in instructional or procedural text (e.g. “putting,” “turn,” “next,” “mixed,” “last”).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Dems

0.75

𝕚

0.74

 Adri

0.73

̞

0.73

ند

0.71

قل

0.71

وه

0.71

Adress

0.71

KH

0.70

 Richards

0.70

POSITIVE LOGITS

 tảo

0.76

 shed

0.70

 conteudo

0.69

 funcionários

0.68

 وعلى

0.68

றிந்த

0.68

netic

0.67

atively

0.66

こと

0.65

 begs

0.65

Activations Density 0.000%