INDEX

Explanations

the other way around

This neuron detects the idiomatic phrase expressing reversal (e.g. “the other way around” or “vice versa”).

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

-1.73

↵↵

-1.53

 That

-1.49

别的

-1.45

おわりに

-1.43

-1.42

 sogen

-1.41

 sufrió

-1.41

 Що

-1.36

POSITIVE LOGITS

givings

2.09

чи

1.80

();

1.68

筑

1.67

++)

1.67

}+

1.66

ynthesis

1.64

'}),

1.63

ankind

1.60

 than

1.58

Activations Density 0.005%