INDEX

Explanations

programming and data structures

This neuron detects expressions of gratitude or praise—words like “thanks,” “thank,” and positive descriptors used in thank-you comments.

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

사를

0.93

사의

0.91

HE

0.90

상

0.87

า

0.87

ласти

0.86

າ

0.84

INA

0.82

이며

0.82

ンの

0.82

POSITIVE LOGITS

 FaceTime

0.93

 rugs

0.89

 BlackBerry

0.88

 Antibodies

0.88

 lobes

0.87

 Flavor

0.87

 TIFF

0.87

 kefir

0.86

 smartphones

0.84

 syrups

0.84

Activations Density 0.009%