INDEX

Explanations

subjects and their states or actions

The neuron selectively activates on personal pronouns (e.g. I, he, she, they) and related simple auxiliary verbs.

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

 også

-1.20

and

-1.17

 myös

-1.04

 ayrıca

-0.97

并

-0.94

 także

-0.93

رر

-0.92

 также

-0.91

 pudieron

-0.88

 Chen

-0.85

POSITIVE LOGITS

ograft

1.05

柢

1.01

 just

1.01

usive

1.00

too

0.98

それを

0.96

 still

0.90

outchouc

0.88

 merely

0.88

 inexist

0.87

Activations Density 0.029%