INDEX

Explanations

scores and ratings

The neuron is looking for numeric tokens—particularly digits and star‐rating numbers.

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

 complained

0.65

 looked

0.63

TOS

0.62

 showed

0.62

NMR

0.62

 Hydrochloride

0.62

लिनी

0.62

लिक

0.61

 sightings

0.61

FCS

0.61

POSITIVE LOGITS

象征

0.73

設け

0.68

因为它

0.67

 града

0.66

本格

0.64

stargo

0.64

最高的

0.64

Rating

0.63

punkte

0.63

 ponieważ

0.63

Activations Density 0.076%