INDEX

Explanations

further refinement, research, communication, decline

The neuron fires on words that signal adding items or enumerating (“more,” “another,” “additional,” etc.).

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

বারেই

1.16

𝙉

1.07

 meisten

1.05

Avec

1.04

Ссылки

1.04

Oku

1.04

ेक्ट

1.03

StartsWith

1.03

NAC

1.03

Renk

1.03

POSITIVE LOGITS

 than

1.66

 niż

1.66

 nữa

1.52

than

1.44

 Than

1.21

 Another

1.19

Than

1.12

↵

1.10

ទៀត

1.07

ier

1.06

Activations Density 2.151%