INDEX

Explanations

disclaimers and key points

The main thing this neuron does is detect strongly positive sentiment or enthusiastic praise (e.g. “absolutely fabulous,” “top-notch,” expressions of love or recommendation).

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

Prevention

0.68

Interpol

0.64

Unified

0.62

Symbol

0.62

Applications

0.61

Template

0.61

ผู้ชม

0.61

Verification

0.60

Internal

0.60

HTTPS

0.60

POSITIVE LOGITS

 maravilh

1.12

 restaurant

1.06

 appartement

1.05

 bungalow

1.05

 motel

1.02

 restoran

0.99

 townhouse

0.97

 eatery

0.94

 ristorante

0.94

 condo

0.94

Activations Density 0.668%