INDEX

Explanations

both ... and

The neuron detects coordinating conjunctions and related function words that join or compare concepts (e.g. “and,” “both,” “while,” “still”).

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

讶

-1.04

总之

-0.97

रि

-0.92

overall

-0.89

 anyone

-0.89

 spørgsmål

-0.88

řed

-0.88

biendo

-0.87

Assassin

-0.86

ÑOS

-0.85

POSITIVE LOGITS

 simultaneously

1.73

 Simultaneously

1.44

同時に

1.38

 zároveň

1.38

 also

1.34

 одновременно

1.33

Also

1.24

AND

1.21

 jednocześnie

1.21

 concurrently

1.19

Activations Density 0.055%