INDEX

Explanations

next week

The neuron strongly activates on words evoking collective struggle, conflict, or major events (e.g. “front lines,” “World War II,” “polarities,” “divisions,” “Congress”).

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

 لأنه

1.02

 kvůli

0.99

แค่

0.93

 ছিলনা

0.91

 díky

0.90

 perché

0.88

 لأن

0.87

不算

0.86

 因為

0.86

 Because

0.85

POSITIVE LOGITS

new

1.06

新的

0.79

 clues

0.78

新たな

0.76

0.73

 developments

0.71

 wondrous

0.70

Activations Density 0.040%