INDEX

Explanations

states of belief or assertion

The neuron specializes in detecting abstract “proposition” words that introduce or label an author’s claims or hypotheses (e.g. assumption, hypothesis, claim, possibility).

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

and

-1.02

oczes

-0.93

얘

-0.83

âce

-0.82

Itu

-0.81

meliharaan

-0.80

 TensorFlow

-0.80

Ꮜ

-0.80

inalis

-0.80

 chose

-0.79

POSITIVE LOGITS

 that

2.89

of

1.69

 rằng

1.13

 bahwa

1.09

 بأن

1.03

že

0.96

 که

0.95

ว่า

0.95

 согласно

0.90

 будто

0.89

Activations Density 0.063%