INDEX

Explanations

historical or present tense markers

The neuron strongly activates on the very first token of a new segment or sentence (i.e. sentence‐initial words).

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

 AUGUST

0.91

 August

0.88

 JULY

0.87

일

0.84

our

0.84

 giữ

0.84

 제거

0.83

Att

0.81

 JANUARY

0.81

 APRIL

0.80

POSITIVE LOGITS

主義

0.75

 possíveis

0.73

 власть

0.72

 problemas

0.71

тся

0.71

 posibles

0.70

вый

0.70

holes

0.70

щей

0.69

 перемен

0.69

Activations Density 0.000%