INDEX

Explanations

until now / since

The neuron flags content‐bearing words—especially nouns and verbs that carry key information (often rarer or more domain-specific terms) rather than common function words.

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 cuando

0.84

 when

0.82

 quando

0.82

 WHEN

0.78

 когда

0.77

when

0.73

 Cuando

0.72

 Когда

0.72

cuando

0.72

khi

0.71

POSITIVE LOGITS

 since

2.45

since

2.36

 منذ

2.31

 depuis

2.28

 sejak

2.13

至今

2.13

 seit

2.11

 bisher

2.09

以來

2.08

 இதுவரை

2.08

Activations Density 0.763%