INDEX

Explanations

again

The neuron fires on the main action words (lexical verbs and phrasal-verb particles) in a sentence.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 länger

-1.22

 таких

-1.20

 ersten

-1.20

 takich

-1.20

 bepaalde

-1.19

 such

-1.18

 erfolgreich

-1.18

 mittler

-1.15

 сча

-1.15

 должны

-1.13

POSITIVE LOGITS

 again

3.52

 more

2.09

Again

1.78

อีก

1.74

again

1.69

 دوباره

1.62

 anymore

1.62

 further

1.53

 other

1.52

 next

1.50

Activations Density 0.034%