INDEX

Explanations

remembered her, convinced Elias, adjusting the

The neuron fires most strongly on verbs—especially action words (often in –ing or simple present forms) describing what someone is doing.

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

she

0.99

彼は

0.97

She

0.91

Ms

0.91

 unserer

0.90

 উনি

0.89

 ésta

0.88

 она

0.88

ihn

0.88

 ellas

0.87

POSITIVE LOGITS

 herself

2.00

 своего

1.38

 తన

1.36

 svého

1.35

 തന്റെ

1.33

 자신의

1.29

 نفسها

1.23

 своему

1.23

 swojego

1.23

 svůj

1.19

Activations Density 0.425%